Which machine library do you use the most?
- Catboost
- Dspy
- gpflow
- Langchain
- LGBM
- LightGBM
- Nixtla
- Nixtla, darts, sktime, gluonts
- Non
- null
- Perpetual
- Pymc
- PyTorch/Lightning/Fast.ai
- Sklearn
- Stable-Baselines3
- TensorFlow
- triton
- XGBoost
Which areas or modality of machine learning / science does your team work on?
Which use-cases does your team use machine learning for?
How long does it take your team to productionise a model (including data ingestion, etc)?
- Less than 3 months
- Less than 6 months
- Less than a month
- Less than a week
- Less than a year
- More than a year
- null
Which cloud platform do you use the most?
- Amazon Web Services
- AWS & GCP
- Azure
- Databricks
- Digital Ocean
- GCP and AWS equally
- Google Cloud Platform
- IBM
- No cloud platform used
- null
- On-prem
Select the top 3 biggest challenges that you face when productionising your machine learning models
For Model Registry and/or Experiment tracking, what tool do you use the most? Skip this question if you don't use any.
- Azure ML Service, Sagemaker
- ClearML
- Comet
- Custom Built In-house tool
- Data Version Control (DVC)
- Domino
- GCS and Clearml
- Google MetaDataStore
- Hopsworks
- MLFlow
- null
- omega-ml
- Optuna
- Sagemaker
- Snowflake
- Spreadsheets
- Vertex AI Experiments
- Weights & Biases
For Feature Store, what tool do you use the most? Skip this question if you don't use any.
- Custom Built In-house tool
- Databricks
- Dataiku
- DynamoDB
- FEAST
- Fennel
- Hopsworks
- None
- null
- selft built / Databricks
- Snowflake
- Tecton
For Vector Database, what tool do you use the most? Skip this question if you don't use any.
- ActiveLoop
- Azure AI Search
- Azure Cognitive Search
- azure's offering in ai search
- Chroma
- ChromaDB
- Custom Built In-house tool
- Databricks
- DynamoDB
- Elastic Search
- Elasticsearch
- Elasticsearch / Opensearch
- Faiss
- gcp
- Hopsworks
- LanceDB
- LanceDB / Turbopuffer
- Milvus
- MongoDB
- NA
- nil
- No vector database
- None
- null
- Opensearch
- opensearch / Azure AI Search
- Opensearch AWS
- opensearch, elasticsearch
- Pgvector
- Pinecone
- PostgreSQL
- Qdrant
- Weaviate
For ETL / Workflow Orchestrator, what tool do you use the most? Skip this question if you don't use any.
- Abinitio
- Airflow
- Amazon Glue
- Argo Workflows
- aws step function
- AWS Step Functions
- Azure data factory, Databricks workflows, azureML pipelines
- Azure devops
- Azure ML pipelines
- Celery
- ClearML
- Custom Built In-house tool
- Dagster
- databricks
- Databricks workflow
- Databricks Workflows
- Dataiku
- Dbt
- Github actions
- Kubeflow
- kubeflow and ray
- kubeflow pipelines
- Meta flow
- Metaflow
- NiFi
- None
- null
- Prefect
- snakemake
- Snowflake
For Model Training and Experimentation Platform, what tool do you use the most? Skip this question if you don't use any.
- Amazon SageMaker
- Anyscale and also Slurm
- Argo Workflows
- Azure ML Studio
- clearml
- Custom Built In-house tool
- Databricks
- Domino
- Google Cloud Vertex AI
- Hopsworks
- kubeflow
- Kubernetes, docker
- Local
- Metaflow
- null
- Omega-ml
- Snowflake
- Weights and biases
For Real Time Model Serving, what tool do you use the most? Skip this question if you don't use any.
- Anyscale/Ray
- AWS lambda
- Azure Managed Online Endpoints
- Azure ML
- BentoML
- Custom Built In-house tool
- Databricks
- FastAPI/Flask Wrapper
- KServe
- Kubernetes
- LMDeploy
- null
- Nvidia Triton Inference Server
- Omega-ml
- Sagemaker
- Seldon Core
- SkyPilot
- Spring Boot
- tensorrt
- TorchServe
- triton
- Triton Inference Server
For Model Monitoring, what tool do you use the most? Skip this question if you don't use any.
- Arize AI
- Azure ML
- Comet
- Custom Built In-house tool
- Domino
- Evidently AI
- Fiddler AI
- Grafana
- Hopsworks
- mlcore
- Model metrics monitoring on Metabase
- NannyML
- Neptune AI
- null
- Omega-ml
- Perpetual ML Suite
- Sagemaker
- WhyLabs
For Central Data Platform / Data Lake, what tool do you use the most? Skip this question if you don't use any.
- AWS / Lakeformation
- Azure DataLake
- Custom Built In-house tool
- Deltalake
- Dremio
- GCP / BigLake
- IBM
- MSSQL
- null
- Snowflake
- TDengine
- Trino
For Managed Foundation Model / LLM Api Services, what tool do you use the most? Skip this question if you don't use any.
- Amazon Bedrock
- Anthropic
- Azure AI
- Custom Built In-house tool
- Ensemble
- fireworks
- Google Gemini
- Mistral
- None
- null
- Omega-ml
- OpenAI
- Snowflake
Which industry are you in?
- Aerospace & defense
- Agriculture
- Automotive
- Biotech
- Consultancy
- consulting
- Data Consulting
- Electronics
- Energy
- Enviroment
- Fashion
- Financial services
- Food
- Food Delivery
- Government
- Healthcare
- Hospitality
- Human Resources
- Insurance
- Manufacturing
- Market Research
- Marketplaces
- Media & Entertainment
- null
- Pharma
- Pharmaceuticals
- Public Sector
- Retail
- Saas
- Technology
- Telecommunications
- Transportation & warehousing
- Travel
- Utilities
How large is your organisation?
- 1,000-5,000 employees
- 10-50 employees
- 250-1,000 employees
- 5,000-50,000 employees
- 50,000+ employees
- 50-250 employees
- Less than 10 employees
- null
How models does your organisation have in production? This can be a guesstimate and doesn't have to be an accurate number.
- 0
- 1
- 10-20
- 100-1000
- 1000+
- 2-5
- 21-100
- 5-9
- null
In 12 months from today, how many models does your organisation plan to have in production by then? This can be a guesstimate and doesn't have to be an accurate number.
- 0
- 1
- 10-20
- 100-1000
- 1000+
- 2-5
- 21-100
- 5-9
- null
Select which of the following your organisation has set up:
What approximate percentage of your models run real-time INFERENCE instead of batch?
- Between 10% and 30%
- Between 50% and 90%
- Less than 10%
- More than 90%
- null
When productionising a machine learning model, does your infrastructure enable any of the following deployment methods?
What is the "job family" of your role?
- Business / Domain Practitioner
- Data Analyst
- Data Engineer
- Data Scientist
- Machine Learning Engineer
- MLOps Engineer
- null
- Product Manager
- Software Engineer
How would you describe your role?
- C-Suite
- Director/VP
- Individual Contributor (Junior to Senior)
- Individual Contributor (Staff+)
- Manager
- null
What is your age?
- 18-21
- 22-24
- 25-29
- 30-34
- 35-39
- 40-44
- 45-49
- 50-54
- 55-59
- null
- Prefer not to share
What is your country of residence?
- Australia
- Austria
- Bangladesh
- Brazil
- Canada
- Colombia
- Czech Republic
- Denmark
- Estonia
- Finland
- France
- Germany
- India
- Israel
- Italy
- Japan
- Luxembourg
- Netherlands
- nl
- Northern Ireland
- Norway
- null
- portugal
- Romania
- Spain
- Switzerland
- Thailand
- The Netherlands
- Turkey
- UAE
- UK
- United Kingdom
- United States
- US
- USA
- Vietnam
How do you self identify?
- Female
- Male
- Non-binary
- null
Loading...
| Which machine library do you use the most? | Which areas or modality of machine learning / science does your team work on? | Which use-cases does your team use machine learning for? | How long does it take your team to productionise a model (including data ingestion, etc)? | Which cloud platform do you use the most? | Select the top 3 biggest challenges that you face when productionising your machine learning models | For Model Registry and/or Experiment tracking, what tool do you use the most? Skip this question if you don't use any. | For Feature Store, what tool do you use the most? Skip this question if you don't use any. | For Vector Database, what tool do you use the most? Skip this question if you don't use any. | For ETL / Workflow Orchestrator, what tool do you use the most? Skip this question if you don't use any. | For Model Training and Experimentation Platform, what tool do you use the most? Skip this question if you don't use any. | For Real Time Model Serving, what tool do you use the most? Skip this question if you don't use any. | For Model Monitoring, what tool do you use the most? Skip this question if you don't use any. | For Central Data Platform / Data Lake, what tool do you use the most? Skip this question if you don't use any. | For Managed Foundation Model / LLM Api Services, what tool do you use the most? Skip this question if you don't use any. | Which industry are you in? | How large is your organisation? | How models does your organisation have in production? This can be a guesstimate and doesn't have to be an accurate number. | In 12 months from today, how many models does your organisation plan to have in production by then? This can be a guesstimate and doesn't have to be an accurate number. | Select which of the following your organisation has set up: | What approximate percentage of your models run real-time INFERENCE instead of batch? | When productionising a machine learning model, does your infrastructure enable any of the following deployment methods? | What is the "job family" of your role? | How would you describe your role? | What is your age? | What is your country of residence? | How do you self identify? |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Sklearn | Causal Inference, Time Series / Forecasting | Fraud, Risk | Less than 6 months | Google Cloud Platform | Machine learning monitoring and observability, Machine learning security, Governance and domain risks | Spreadsheets | Hopsworks | null | Airflow | Databricks | FastAPI/Flask Wrapper | null | Deltalake | OpenAI | Retail | 5,000-50,000 employees | 100-1000 | 1000+ | A Developer Productivity Team which also covers machine learning, Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Risk & Governance Function, AI Inventory (Keeping track of all AI usecases and models) | Between 10% and 30% | CI/CD for continuous deployment | Software Engineer | Director/VP | 30-34 | Germany | Male |
| LightGBM | Text / NLP (Non-LLM), Regression | Demand Forecasting, Search, Recommender systems, Pricing | Less than 3 months | Amazon Web Services | Inconsistency of training and experimentation environments, Building production-grade machine learning and data pipelines | Weights & Biases | null | null | Airflow | Custom Built In-house tool | FastAPI/Flask Wrapper | null | AWS / Lakeformation | null | Technology | 50-250 employees | 10-20 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | More than 90% | Canary Deployments, Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | Data Scientist | Manager | 35-39 | null | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision | Autonomous driving | Less than 3 months | On-prem | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation, Access to relevant data for training | Comet | Custom Built In-house tool | null | Argo Workflows | Argo Workflows | null | Comet | Custom Built In-house tool | null | Transportation & warehousing | 1,000-5,000 employees | 1000+ | 1000+ | A Developer Productivity Team which also covers machine learning, Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | More than 90% | Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Staff+) | 45-49 | null | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision, LLMs | Search | Less than 3 months | Amazon Web Services | Access to relevant data for training, Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | Spreadsheets | null | Weaviate | Custom Built In-house tool | null | LMDeploy | null | null | OpenAI | Fashion | 50-250 employees | 2-5 | 5-9 | null | Between 50% and 90% | null | MLOps Engineer | Individual Contributor (Junior to Senior) | 25-29 | Bangladesh | Male |
| Sklearn | LLMs, Tabular | Fraud | Less than 3 months | Amazon Web Services | Inconsistency of training and experimentation environments, Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | MLFlow | null | null | AWS Step Functions | Amazon SageMaker | FastAPI/Flask Wrapper | null | AWS / Lakeformation | Mistral | Technology | 250-1,000 employees | 5-9 | 10-20 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | More than 90% | Canary Deployments, Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Junior to Senior) | 35-39 | Canada | Male |
| PyTorch/Lightning/Fast.ai | LLMs | Marketing Intelligence | Less than a week | Google Cloud Platform | Machine learning security, Governance and domain risks | Custom Built In-house tool | Hopsworks | Pinecone | null | null | null | null | null | null | Technology | Less than 10 employees | null | null | null | null | null | null | null | null | null | null |
| Catboost | LLMs, Recommender Systems, Tabular, Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Recommender systems, Risk, Pricing | Less than 3 months | Amazon Web Services | Access to relevant data for training | MLFlow | Custom Built In-house tool | Pinecone | Airflow | Custom Built In-house tool | FastAPI/Flask Wrapper | Custom Built In-house tool | Snowflake | OpenAI | Retail | 250-1,000 employees | 10-20 | 21-100 | Data Platform / Data Engineering Organisation | Between 10% and 30% | Progressive Rollouts, CI/CD for continuous deployment | Data Scientist | Individual Contributor (Staff+) | 40-44 | Romania | Male |
| Sklearn | Time Series / Forecasting | null | Less than 3 months | Amazon Web Services | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation | Spreadsheets | null | ActiveLoop | Airflow | Custom Built In-house tool | Custom Built In-house tool | Arize AI | AWS / Lakeformation | OpenAI | Technology | 10-50 employees | 5-9 | 10-20 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | More than 90% | Canary Deployments, A/B Tests for Models, Progressive Rollouts | Machine Learning Engineer | C-Suite | 25-29 | USA | Male |
| TensorFlow | Image / Computer Vision, Recommender Systems, Text / NLP (Non-LLM) | Marketing Intelligence, Search, Recommender systems | Less than a week | No cloud platform used | Showcasing business impact and business value, Machine learning security, Governance and domain risks | MLFlow | None | None | Airflow | Custom Built In-house tool | FastAPI/Flask Wrapper | Custom Built In-house tool | Snowflake | OpenAI | Data Consulting | Less than 10 employees | 10-20 | 21-100 | A Developer Productivity Team which also covers machine learning | Less than 10% | A/B Tests for Models, CI/CD for continuous deployment | Data Scientist | Individual Contributor (Junior to Senior) | 30-34 | USA | Male |
| PyTorch/Lightning/Fast.ai | Recommender Systems, Search, Text / NLP (Non-LLM) | Recommender systems | More than a year | Google Cloud Platform | Gaps in tooling and support for model productionisation, Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | null | null | Faiss | null | Google Cloud Vertex AI | FastAPI/Flask Wrapper | null | GCP / BigLake | OpenAI | Media & Entertainment | 50-250 employees | 100-1000 | 21-100 | Central Machine Learning Platform / Team | Between 10% and 30% | A/B Tests for Models, CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Manager | 30-34 | Japan | Male |
| PyTorch/Lightning/Fast.ai | Time Series / Forecasting | Demand Forecasting | Less than 6 months | Amazon Web Services | Lack of specialised engineers, Showcasing business impact and business value, Machine learning monitoring and observability | MLFlow | null | null | Airflow | Databricks | null | null | Deltalake | null | Retail | 5,000-50,000 employees | null | null | Central Machine Learning Platform / Team, AI Risk & Governance Function, AI Inventory (Keeping track of all AI usecases and models) | Less than 10% | Canary Deployments, CI/CD for continuous deployment | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 30-34 | France | Male |
| Sklearn | LLMs, Text / NLP (Non-LLM), Time Series / Forecasting | Search, Recommender systems, Risk, Healthcare | Less than 6 months | Azure | Inconsistency of training and experimentation environments, Access to relevant data for training, Lack of specialised engineers | Custom Built In-house tool | Custom Built In-house tool | null | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Azure DataLake | OpenAI | Healthcare | 5,000-50,000 employees | 21-100 | 100-1000 | A Developer Productivity Team which also covers machine learning, Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Less than 10% | Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Staff+) | 40-44 | USA | Male |
| Sklearn | Image / Computer Vision, Recommender Systems | Recommender systems, Visual Detection in Video | Less than 6 months | Azure | Gaps in tooling and support for model productionisation, Building production-grade machine learning and data pipelines | null | null | Custom Built In-house tool | null | null | null | null | null | OpenAI | Agriculture | 1,000-5,000 employees | 2-5 | 10-20 | null | null | Progressive Rollouts, Development-Staging-Production Environments | Business / Domain Practitioner | Individual Contributor (Staff+) | 50-54 | USA | Male |
| PyTorch/Lightning/Fast.ai | Recommender Systems | Marketing Intelligence | More than a year | Amazon Web Services | Lack of specialised engineers, Lack of specialised data scientists | MLFlow | Custom Built In-house tool | null | Airflow | Amazon SageMaker | Custom Built In-house tool | Custom Built In-house tool | AWS / Lakeformation | null | Technology | 50-250 employees | 100-1000 | 100-1000 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | More than 90% | Canary Deployments, A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment | null | Individual Contributor (Staff+) | 30-34 | Israel | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision | null | Less than a week | Amazon Web Services | Access to relevant data for training, Showcasing business impact and business value, Building production-grade machine learning and data pipelines | Data Version Control (DVC) | null | null | null | null | null | null | null | null | Food | 10-50 employees | 5-9 | 5-9 | AI Inventory (Keeping track of all AI usecases and models) | Less than 10% | CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Staff+) | 30-34 | null | Male |
| PyTorch/Lightning/Fast.ai | LLMs, Recommender Systems | Search, Recommender systems | Less than 3 months | Amazon Web Services | Gaps in tooling and support for model productionisation, Machine learning security, Governance and domain risks | Data Version Control (DVC) | Custom Built In-house tool | Milvus | Argo Workflows | Amazon SageMaker | Sagemaker | Neptune AI | AWS / Lakeformation | Azure AI | Telecommunications | 50,000+ employees | 100-1000 | 1000+ | A Developer Productivity Team which also covers machine learning, Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 50% and 90% | Canary Deployments, A/B Tests for Models, CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Staff+) | 35-39 | Germany | Male |
| Sklearn | Tabular | Fraud | More than a year | Amazon Web Services | Inconsistency of training and experimentation environments, Lack of specialised engineers | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Amazon Glue | Amazon SageMaker | Sagemaker | Custom Built In-house tool | AWS / Lakeformation | Custom Built In-house tool | Financial services | 50,000+ employees | 1000+ | 100-1000 | Central Machine Learning Platform / Team | Between 10% and 30% | CI/CD for continuous deployment | Machine Learning Engineer | Manager | 30-34 | null | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision | Pharmaceuticals | Less than 6 months | Google Cloud Platform | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation, Machine learning monitoring and observability | Weights & Biases | null | null | Prefect | Anyscale and also Slurm | FastAPI/Flask Wrapper | null | GCP / BigLake | null | Biotech | 250-1,000 employees | 10-20 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Inventory (Keeping track of all AI usecases and models) | Less than 10% | CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Junior to Senior) | Prefer not to share | USA | Female |
| Sklearn | Image / Computer Vision, LLMs, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Pricing | Less than a month | Azure | Gaps in tooling and support for model productionisation, Lack of specialised engineers, Building production-grade machine learning and data pipelines | MLFlow | null | Azure AI Search | null | Databricks | FastAPI/Flask Wrapper | null | Azure DataLake | Azure AI | Technology | 250-1,000 employees | 2-5 | 5-9 | A Developer Productivity Team which also covers machine learning, Data Platform / Data Engineering Organisation | Between 10% and 30% | Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 25-29 | Brazil | Male |
| Sklearn | Tabular | Demand Forecasting, Marketing Intelligence, Risk, Pricing | Less than a week | Azure | Gaps in tooling and support for model productionisation, Lack of specialised engineers, Showcasing business impact and business value | MLFlow | Custom Built In-house tool | nil | databricks | Databricks | Databricks | mlcore | Deltalake | Azure AI | consulting | 1,000-5,000 employees | 10-20 | 10-20 | A Developer Productivity Team which also covers machine learning, Central Machine Learning Platform / Team | Less than 10% | CI/CD for continuous deployment | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 30-34 | India | Male |
| Sklearn | Image / Computer Vision, LLMs, Recommender Systems, Search, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Search, Recommender systems | Less than 6 months | Amazon Web Services | Lack of specialised engineers, Machine learning monitoring and observability, Machine learning security | Weights & Biases | Hopsworks | Weaviate | Airflow | Amazon SageMaker | BentoML | Evidently AI | AWS / Lakeformation | OpenAI | Aerospace & defense | Less than 10 employees | 1 | null | null | null | A/B Tests for Models, CI/CD for continuous deployment, Development-Staging-Production Environments | Data Scientist | Individual Contributor (Junior to Senior) | 18-21 | India | Male |
| PyTorch/Lightning/Fast.ai | LLMs, Recommender Systems, Search, Text / NLP (Non-LLM) | Search, Property Prediction (Molecular) | Less than 3 months | Amazon Web Services | Building production-grade machine learning and data pipelines, Governance and domain risks | Custom Built In-house tool | Custom Built In-house tool | Milvus | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Technology | 50,000+ employees | 1000+ | 1000+ | A Developer Productivity Team which also covers machine learning, Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Risk & Governance Function | Between 10% and 30% | Canary Deployments, A/B Tests for Models, CI/CD for continuous deployment | Software Engineer | Director/VP | 45-49 | USA | Male |
| Sklearn | LLMs, Text / NLP (Non-LLM), Time Series / Forecasting | Marketing Intelligence, Risk, Pricing | Less than a year | Amazon Web Services | null | null | null | null | null | null | FastAPI/Flask Wrapper | null | AWS / Lakeformation | null | Electronics | 50-250 employees | 21-100 | 21-100 | A Developer Productivity Team which also covers machine learning, AI Risk & Governance Function, AI Inventory (Keeping track of all AI usecases and models) | More than 90% | null | MLOps Engineer | C-Suite | 35-39 | null | Male |
| Dspy | LLMs, Recommender Systems, Search, Text / NLP (Non-LLM) | Marketing Intelligence, Recommender systems | null | Amazon Web Services | Access to relevant data for training, Showcasing business impact and business value, Machine learning monitoring and observability | MLFlow | null | null | Airflow | null | FastAPI/Flask Wrapper | null | null | OpenAI | null | 250-1,000 employees | null | null | Data Platform / Data Engineering Organisation | null | CI/CD for continuous deployment | Machine Learning Engineer | Individual Contributor (Junior to Senior) | null | null | null |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision, LLMs, Tabular, Time Series / Forecasting | defect detection | Less than 6 months | Azure | Lack of specialised engineers, Machine learning monitoring and observability, Governance and domain risks | Weights & Biases | Dataiku | null | Dataiku | Azure ML Studio | Triton Inference Server | Custom Built In-house tool | Snowflake | Azure AI | Technology | 5,000-50,000 employees | 10-20 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Inventory (Keeping track of all AI usecases and models) | Between 50% and 90% | Development-Staging-Production Environments | Data Scientist | Individual Contributor (Junior to Senior) | 30-34 | Germany | Male |
| XGBoost | Text / NLP (Non-LLM) | Risk | More than a year | Amazon Web Services | Inconsistency of training and experimentation environments, Access to relevant data for training, Governance and domain risks | MLFlow | FEAST | Qdrant | Dbt | Amazon SageMaker | null | WhyLabs | AWS / Lakeformation | null | Healthcare | 10-50 employees | 0 | 2-5 | null | null | CI/CD for continuous deployment | Data Scientist | Manager | null | null | null |
| Catboost | LLMs, Recommender Systems, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Marketing Intelligence, Recommender systems, Fraud, Risk | Less than 6 months | Azure | Access to relevant data for training, Showcasing business impact and business value | MLFlow | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Databricks | FastAPI/Flask Wrapper | NannyML | Deltalake | Azure AI | Telecommunications | 5,000-50,000 employees | 10-20 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Risk & Governance Function | Less than 10% | A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | Data Scientist | Individual Contributor (Staff+) | 25-29 | Thailand | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision, Search, Text / NLP (Non-LLM) | Search | Less than a month | Amazon Web Services | Lack of specialised engineers, Building production-grade machine learning and data pipelines | MLFlow | null | Milvus | Argo Workflows | Amazon SageMaker | Nvidia Triton Inference Server | null | null | null | Retail | 50-250 employees | 2-5 | 5-9 | null | Between 50% and 90% | null | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 35-39 | Turkey | Male |
| Stable-Baselines3 | RL | Optimization | Less than a month | Amazon Web Services | Inconsistency of training and experimentation environments, Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | Custom Built In-house tool | Custom Built In-house tool | null | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | null | null | Energy | Less than 10 employees | 0 | 1 | Central Machine Learning Platform / Team | More than 90% | null | Data Scientist | Individual Contributor (Staff+) | 25-29 | USA | Male |
| PyTorch/Lightning/Fast.ai | LLMs, Search, Text / NLP (Non-LLM) | Search | Less than 3 months | Google Cloud Platform | Inconsistency of training and experimentation environments, Machine learning monitoring and observability, Machine learning security | MLFlow | null | Milvus | Argo Workflows | Custom Built In-house tool | Custom Built In-house tool | null | null | OpenAI | Technology | 50-250 employees | 5-9 | 10-20 | Central Machine Learning Platform / Team | More than 90% | Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 25-29 | USA | Male |
| TensorFlow | Tabular | Marketing Intelligence, Pricing | Less than 3 months | Amazon Web Services | Gaps in tooling and support for model productionisation, Building production-grade machine learning and data pipelines | Weights & Biases | null | LanceDB | Metaflow | Custom Built In-house tool | FastAPI/Flask Wrapper | Evidently AI | null | OpenAI | Financial services | 1,000-5,000 employees | 1000+ | 1000+ | Data Platform / Data Engineering Organisation | Between 10% and 30% | null | Machine Learning Engineer | Individual Contributor (Staff+) | 30-34 | USA | Male |
| PyTorch/Lightning/Fast.ai | LLMs, Text / NLP (Non-LLM) | Recommender systems, Agentic systems, agents assists | Less than a month | Amazon Web Services | Inconsistency of training and experimentation environments, Access to relevant data for training, Showcasing business impact and business value | Custom Built In-house tool | null | Milvus | Custom Built In-house tool | null | FastAPI/Flask Wrapper | null | Snowflake | Azure AI | Telecommunications | 250-1,000 employees | 2-5 | 10-20 | Central Machine Learning Platform / Team | More than 90% | Development-Staging-Production Environments | Product Manager | Director/VP | 40-44 | null | Male |
| PyTorch/Lightning/Fast.ai | LLMs, Text / NLP (Non-LLM) | Fraud | Less than a year | Google Cloud Platform | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation, Access to relevant data for training | Weights & Biases | FEAST | Milvus | Argo Workflows | Google Cloud Vertex AI | FastAPI/Flask Wrapper | Neptune AI | Snowflake | OpenAI | Financial services | 50-250 employees | 21-100 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 50% and 90% | A/B Tests for Models | Software Engineer | Manager | 25-29 | Vietnam | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision | Fraud | Less than a month | Google Cloud Platform | Inconsistency of training and experimentation environments, Showcasing business impact and business value, Governance and domain risks | ClearML | Custom Built In-house tool | null | Custom Built In-house tool | null | Custom Built In-house tool | null | null | null | Retail | 10-50 employees | 10-20 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | More than 90% | A/B Tests for Models, CI/CD for continuous deployment | Software Engineer | Individual Contributor (Staff+) | 40-44 | Germany | Male |
| Sklearn | LLMs, Tabular | Fraud | Less than a year | Amazon Web Services | Gaps in tooling and support for model productionisation, Access to relevant data for training, Machine learning monitoring and observability | Spreadsheets | null | null | null | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | OpenAI | Financial services | 250-1,000 employees | 21-100 | 21-100 | Data Platform / Data Engineering Organisation | More than 90% | Development-Staging-Production Environments | Data Scientist | Manager | 35-39 | Spain | Male |
| Sklearn | Time Series / Forecasting | Risk | Less than 6 months | Azure | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation, Showcasing business impact and business value | MLFlow | null | null | null | Azure ML Studio | FastAPI/Flask Wrapper | Azure ML | Deltalake | null | Energy | 5,000-50,000 employees | 100-1000 | 1000+ | Data Platform / Data Engineering Organisation | Less than 10% | CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Staff+) | 35-39 | Netherlands | Male |
| Catboost | LLMs, Tabular | Marketing Intelligence, Support automation | Less than 3 months | Amazon Web Services | Lack of specialised engineers, Showcasing business impact and business value, Machine learning monitoring and observability | ClearML | null | null | Airflow | clearml | FastAPI/Flask Wrapper | null | null | OpenAI | Media & Entertainment | 250-1,000 employees | 2-5 | 5-9 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 10% and 30% | Development-Staging-Production Environments | Machine Learning Engineer | Manager | 35-39 | Estonia | Male |
| PyTorch/Lightning/Fast.ai | Recommender Systems, Search | Search, Recommender systems | null | Google Cloud Platform | Lack of specialised engineers, Building production-grade machine learning and data pipelines | Vertex AI Experiments | null | null | Airflow | Google Cloud Vertex AI | null | null | GCP / BigLake | null | Travel | 250-1,000 employees | 10-20 | 21-100 | Data Platform / Data Engineering Organisation | Less than 10% | null | Data Scientist | Individual Contributor (Junior to Senior) | null | null | Male |
| TensorFlow | Causal Inference, Search, Time Series / Forecasting | Demand Forecasting, Search, Fraud, Risk | Less than 3 months | Azure | Access to relevant data for training, Showcasing business impact and business value, Building production-grade machine learning and data pipelines | MLFlow | null | null | Prefect | Databricks | Databricks | null | Deltalake | OpenAI | Financial services | Less than 10 employees | 1 | 2-5 | A Developer Productivity Team which also covers machine learning | Less than 10% | CI/CD for continuous deployment, Development-Staging-Production Environments | Data Engineer | Individual Contributor (Staff+) | 30-34 | USA | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision, LLMs | Search, Recommender systems | Less than a week | Digital Ocean | Showcasing business impact and business value, Building production-grade machine learning and data pipelines, Machine learning security | Data Version Control (DVC) | null | Pgvector | Prefect | Custom Built In-house tool | SkyPilot | Evidently AI | Custom Built In-house tool | Custom Built In-house tool | Technology | Less than 10 employees | 2-5 | 5-9 | A Developer Productivity Team which also covers machine learning, Central Machine Learning Platform / Team | Between 50% and 90% | A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment | Machine Learning Engineer | Individual Contributor (Junior to Senior) | Prefer not to share | USA | Male |
| PyTorch/Lightning/Fast.ai | Recommender Systems, Search, Text / NLP (Non-LLM) | Search, Recommender systems | Less than 6 months | Amazon Web Services | Lack of specialised engineers, Showcasing business impact and business value, Building production-grade machine learning and data pipelines | MLFlow | null | Opensearch | Argo Workflows | null | FastAPI/Flask Wrapper | null | null | Anthropic | Marketplaces | 250-1,000 employees | 1 | 2-5 | A Developer Productivity Team which also covers machine learning | More than 90% | A/B Tests for Models, Progressive Rollouts, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 25-29 | Austria | null |
| PyTorch/Lightning/Fast.ai | LLMs, Search, Text / NLP (Non-LLM) | Search | Less than a week | Amazon Web Services | Gaps in tooling and support for model productionisation, Lack of specialised engineers, Building production-grade machine learning and data pipelines | MLFlow | Custom Built In-house tool | null | Airflow | Amazon SageMaker | BentoML | Custom Built In-house tool | Deltalake | null | Technology | 250-1,000 employees | 21-100 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 50% and 90% | Canary Deployments, Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Manager | 35-39 | USA | Male |
| XGBoost | LLMs, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Pricing | Less than 6 months | Azure | Showcasing business impact and business value, Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | MLFlow | selft built / Databricks | null | Airflow | Databricks | FastAPI/Flask Wrapper | Custom Built In-house tool | Snowflake | Azure AI | Energy | 1,000-5,000 employees | 1000+ | 1000+ | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 10% and 30% | Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Manager | 40-44 | Netherlands | Male |
| Sklearn | Causal Inference, Tabular, Time Series / Forecasting | Synthetic data generation | Less than 3 months | Amazon Web Services | Lack of specialised engineers, Lack of specialised data scientists | Custom Built In-house tool | null | null | Argo Workflows | Custom Built In-house tool | null | null | null | null | Technology | 10-50 employees | 2-5 | 5-9 | Data Platform / Data Engineering Organisation | Less than 10% | A/B Tests for Models, CI/CD for continuous deployment | Product Manager | Manager | 30-34 | USA | Female |
| Perpetual | LLMs, Tabular, Time Series / Forecasting | Fraud, Risk | Less than 3 months | Amazon Web Services | Gaps in tooling and support for model productionisation, Lack of specialised data scientists, Showcasing business impact and business value | null | null | azure ai search | null | Databricks | FastAPI/Flask Wrapper | Perpetual ML Suite | Snowflake | Azure AI | Financial services | 1,000-5,000 employees | 10-20 | 21-100 | null | Between 10% and 30% | null | Machine Learning Engineer | Individual Contributor (Staff+) | 35-39 | Turkey | Male |
| Sklearn | LLMs, Recommender Systems, Tabular, Text / NLP (Non-LLM) | Demand Forecasting, Recommender systems, Fraud, Risk, Pricing | Less than 3 months | Google Cloud Platform | Showcasing business impact and business value, Machine learning monitoring and observability | MLFlow | null | null | Airflow | Custom Built In-house tool | FastAPI/Flask Wrapper | Evidently AI | GCP / BigLake | OpenAI | Technology | 250-1,000 employees | 21-100 | 21-100 | Central Machine Learning Platform / Team | Less than 10% | Canary Deployments | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 35-39 | null | null |
| PyTorch/Lightning/Fast.ai | LLMs | Marketing Intelligence, Strategy | Less than a month | Google Cloud Platform | Inconsistency of training and experimentation environments, Machine learning monitoring and observability, Machine learning security | Custom Built In-house tool | Custom Built In-house tool | Pinecone | NiFi | Custom Built In-house tool | FastAPI/Flask Wrapper | null | Custom Built In-house tool | Anthropic | Technology | Less than 10 employees | 2-5 | 2-5 | A Developer Productivity Team which also covers machine learning, Data Platform / Data Engineering Organisation, AI Inventory (Keeping track of all AI usecases and models) | Less than 10% | null | null | Director/VP | null | null | null |
| Sklearn | Search, Tabular, Time Series / Forecasting | Demand Forecasting, Fraud, Risk | Less than 3 months | Amazon Web Services | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation | MLFlow | null | null | Airflow | Databricks | FastAPI/Flask Wrapper | Custom Built In-house tool | Snowflake | Custom Built In-house tool | Technology | 50,000+ employees | 1000+ | 1000+ | Data Platform / Data Engineering Organisation, AI Risk & Governance Function | Less than 10% | Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Staff+) | 25-29 | India | Male |
| Sklearn | Tabular, Time Series / Forecasting | Risk, Pricing | Less than a month | Amazon Web Services | Inconsistency of training and experimentation environments, Access to relevant data for training | MLFlow | Custom Built In-house tool | null | Airflow | Custom Built In-house tool | FastAPI/Flask Wrapper | Custom Built In-house tool | null | null | Financial services | 50-250 employees | 5-9 | 5-9 | Central Machine Learning Platform / Team | Less than 10% | CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Junior to Senior) | 45-49 | Colombia | Male |
| PyTorch/Lightning/Fast.ai | LLMs | Search, Recommender systems | Less than 3 months | Amazon Web Services | Lack of specialised engineers, Lack of specialised data scientists, Showcasing business impact and business value | MLFlow | Databricks | Databricks | Databricks Workflows | Databricks | Databricks | null | null | null | Technology | 250-1,000 employees | 21-100 | 100-1000 | null | null | null | null | null | null | null | null |
| XGBoost | LLMs, Recommender Systems, Time Series / Forecasting | Demand Forecasting, Fraud, Risk | Less than a week | Amazon Web Services | Showcasing business impact and business value | Snowflake | Snowflake | null | Snowflake | Snowflake | Custom Built In-house tool | null | Snowflake | Snowflake | Transportation & warehousing | 5,000-50,000 employees | 10-20 | 21-100 | Data Platform / Data Engineering Organisation | Less than 10% | Canary Deployments, A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | Data Scientist | Individual Contributor (Staff+) | 35-39 | Germany | Male |
| PyTorch/Lightning/Fast.ai | Causal Inference, Image / Computer Vision, LLMs, Recommender Systems, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Recommender systems, Risk, Climate Modelling, Geospatial AI | Less than a month | Amazon Web Services | Gaps in tooling and support for model productionisation, Access to relevant data for training, Governance and domain risks | Custom Built In-house tool | Custom Built In-house tool | null | Airflow | Databricks | Databricks | Fiddler AI | AWS / Lakeformation | Custom Built In-house tool | Insurance | 50,000+ employees | 21-100 | 100-1000 | Data Platform / Data Engineering Organisation, AI Risk & Governance Function, AI Inventory (Keeping track of all AI usecases and models) | Less than 10% | CI/CD for continuous deployment, Development-Staging-Production Environments | Business / Domain Practitioner | Director/VP | 45-49 | United Kingdom | Male |
| Nixtla | Time Series / Forecasting | Demand Forecasting | Less than a month | Amazon Web Services | Inconsistency of training and experimentation environments, Lack of specialised engineers | MLFlow | null | null | null | Databricks | Custom Built In-house tool | null | Deltalake | Amazon Bedrock | Retail | 50,000+ employees | 21-100 | 100-1000 | Data Platform / Data Engineering Organisation | Between 10% and 30% | A/B Tests for Models, CI/CD for continuous deployment, Development-Staging-Production Environments | Data Scientist | Individual Contributor (Junior to Senior) | 30-34 | France | Male |
| Sklearn | Causal Inference, Tabular, Time Series / Forecasting | Demand Forecasting, Recommender systems, Predictive maintenance | Less than 3 months | Azure | Inconsistency of training and experimentation environments, Machine learning monitoring and observability | MLFlow | null | null | Azure data factory, Databricks workflows, azureML pipelines | Azure ML Studio | null | null | Deltalake | Azure AI | Energy | 5,000-50,000 employees | 5-9 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Risk & Governance Function | Less than 10% | Canary Deployments, A/B Tests for Models, CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Staff+) | 35-39 | Germany | Male |
| TensorFlow | Image / Computer Vision | Image Classification | Less than 3 months | Amazon Web Services | Inconsistency of training and experimentation environments, Access to relevant data for training, Building production-grade machine learning and data pipelines | ClearML | null | null | Airflow | Amazon SageMaker | FastAPI/Flask Wrapper | null | AWS / Lakeformation | null | Utilities | 50-250 employees | 2-5 | 5-9 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 10% and 30% | Canary Deployments, A/B Tests for Models, Progressive Rollouts, Development-Staging-Production Environments | Data Scientist | Manager | 40-44 | Israel | null |
| TensorFlow | Causal Inference, LLMs, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Fraud, Risk, Pricing | More than a year | Amazon Web Services | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation, Machine learning monitoring and observability | Spreadsheets | null | null | Airflow | Amazon SageMaker | Sagemaker | null | Custom Built In-house tool | OpenAI | Financial services | 50-250 employees | 5-9 | 10-20 | A Developer Productivity Team which also covers machine learning | null | Canary Deployments, A/B Tests for Models, Progressive Rollouts | Product Manager | Individual Contributor (Staff+) | null | null | null |
| XGBoost | Tabular | Pricing | Less than 3 months | Azure | Inconsistency of training and experimentation environments, Access to relevant data for training | MLFlow | null | null | Databricks workflows | Databricks | FastAPI/Flask Wrapper | Custom Built In-house tool | Azure DataLake | null | Financial services | 50-250 employees | 2-5 | 5-9 | Data Platform / Data Engineering Organisation | More than 90% | CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Staff+) | 30-34 | null | Male |
| Sklearn | Causal Inference, Image / Computer Vision, Recommender Systems, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Recommender systems, Fraud, Risk, Pricing | Less than 6 months | Amazon Web Services | Inconsistency of training and experimentation environments, Access to relevant data for training, Lack of specialised engineers | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Amazon SageMaker | Sagemaker | Custom Built In-house tool | AWS / Lakeformation | Amazon Bedrock | Financial services | 50,000+ employees | 100-1000 | 21-100 | null | Between 10% and 30% | CI/CD for continuous deployment | Machine Learning Engineer | Manager | 30-34 | null | Male |
| PyTorch/Lightning/Fast.ai | LLMs | null | Less than 3 months | Amazon Web Services | Access to relevant data for training, Showcasing business impact and business value, Machine learning monitoring and observability | null | Custom Built In-house tool | Elastic Search | null | Custom Built In-house tool | null | null | null | OpenAI | Media & Entertainment | 1,000-5,000 employees | 2-5 | null | null | null | null | null | null | null | null | null |
| Sklearn | LLMs, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Search, Pricing | Less than 6 months | Amazon Web Services | Lack of specialised engineers, Showcasing business impact and business value, Machine learning monitoring and observability | MLFlow | Custom Built In-house tool | Pinecone | Kubeflow | Custom Built In-house tool | FastAPI/Flask Wrapper | Custom Built In-house tool | Snowflake | Custom Built In-house tool | Transportation & warehousing | 50,000+ employees | 21-100 | 21-100 | Central Machine Learning Platform / Team | More than 90% | Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Staff+) | Prefer not to share | Germany | null |
| null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
| triton | Tabular, Time Series / Forecasting | Fraud | Less than a month | Amazon Web Services | Gaps in tooling and support for model productionisation, Machine learning monitoring and observability, Governance and domain risks | Custom Built In-house tool | Custom Built In-house tool | null | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | null | Custom Built In-house tool | null | Financial services | 250-1,000 employees | 21-100 | 100-1000 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Risk & Governance Function | More than 90% | Canary Deployments, A/B Tests for Models, Progressive Rollouts | MLOps Engineer | Manager | null | null | null |
| Sklearn | Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Recommender systems | Less than 3 months | Azure | Inconsistency of training and experimentation environments, Lack of specialised engineers, Lack of specialised data scientists | Data Version Control (DVC) | FEAST | Weaviate | Airflow | Azure ML Studio | Databricks | Evidently AI | Deltalake | Azure AI | Technology | 50-250 employees | 2-5 | 10-20 | Central Machine Learning Platform / Team | Between 10% and 30% | CI/CD for continuous deployment | Data Scientist | Manager | 30-34 | null | Male |
| LGBM | Causal Inference, Recommender Systems, Time Series / Forecasting | Search, Recommender systems, Fraud, Risk, Pricing | Less than 3 months | Amazon Web Services | Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | Spreadsheets | Custom Built In-house tool | null | Argo Workflows | Custom Built In-house tool | FastAPI/Flask Wrapper | null | Trino | null | Food | 1,000-5,000 employees | 21-100 | 10-20 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 50% and 90% | Canary Deployments, A/B Tests for Models, CI/CD for continuous deployment | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 40-44 | Spain | null |
| null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
| lightgbm | Recommender Systems | Recommender systems | Less than 3 months | Azure | Access to relevant data for training, Showcasing business impact and business value, Machine learning monitoring and observability | MLFlow | Databricks | Databricks | Databricks | Databricks | Databricks | Custom Built In-house tool | Deltalake | Azure AI | Retail | 50,000+ employees | 1000+ | 1000+ | A Developer Productivity Team which also covers machine learning | Between 50% and 90% | A/B Tests for Models, CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Manager | 30-34 | Netherlands | Female |
| TensorFlow | Image / Computer Vision, LLMs, Recommender Systems | Recommender systems, Computer Vision | Less than 6 months | Amazon Web Services | Access to relevant data for training, Lack of specialised engineers, Governance and domain risks | MLFlow | null | Opensearch AWS | None | Amazon SageMaker | Sagemaker | Grafana | AWS / Lakeformation | OpenAI | Media & Entertainment | 1,000-5,000 employees | 2-5 | 10-20 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Less than 10% | A/B Tests for Models, CI/CD for continuous deployment | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 30-34 | Switzerland | Male |
| TensorFlow | Causal Inference, Image / Computer Vision, LLMs, Recommender Systems, Search, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Search, Recommender systems, Pricing | Less than a week | Amazon Web Services | Showcasing business impact and business value, Machine learning monitoring and observability, Governance and domain risks | MLFlow | Snowflake | DynamoDB | Airflow | Amazon SageMaker | Sagemaker | Sagemaker | AWS / Lakeformation | Azure AI | Media & Entertainment | 5,000-50,000 employees | 100-1000 | 1000+ | Data Platform / Data Engineering Organisation, AI Risk & Governance Function | Between 50% and 90% | Canary Deployments, CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Staff+) | 35-39 | Netherlands | Male |
| Sklearn | Tabular, Time Series / Forecasting | Demand Forecasting, Anomaly detection | Less than a month | Azure | Access to relevant data for training, Lack of specialised engineers, Showcasing business impact and business value | MLFlow | null | null | null | Databricks | Custom Built In-house tool | null | Azure DataLake | null | Energy | 10-50 employees | 2-5 | 5-9 | Data Platform / Data Engineering Organisation | Less than 10% | CI/CD for continuous deployment, Development-Staging-Production Environments | Software Engineer | Manager | 45-49 | Norway | Male |
| XGBoost | Recommender Systems, Tabular, Time Series / Forecasting | Recommender systems, Fraud, Risk, Pricing | Less than 6 months | Amazon Web Services | Access to relevant data for training, Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | null | null | null | null | Amazon SageMaker | FastAPI/Flask Wrapper | null | null | null | Telecommunications | Less than 10 employees | 2-5 | 10-20 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 10% and 30% | Development-Staging-Production Environments | Business / Domain Practitioner | Director/VP | 50-54 | Netherlands | Male |
| Sklearn | Time Series / Forecasting | Demand Forecasting | Less than 6 months | Azure | Machine learning monitoring and observability | MLFlow | null | null | Azure devops | Azure ML Studio | null | Azure ml | Deltalake | null | Utilities | 50,000+ employees | 2-5 | 5-9 | Central Machine Learning Platform / Team | Less than 10% | CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Junior to Senior) | 30-34 | United Kingdom | Male |
| Sklearn | Causal Inference, Recommender Systems, Time Series / Forecasting | Demand Forecasting, Recommender systems, Fraud, Risk | Less than a week | Amazon Web Services | Gaps in tooling and support for model productionisation, Showcasing business impact and business value, Governance and domain risks | Domino | null | Pinecone | Airflow | Domino | Seldon Core | Domino | Snowflake | Ensemble | Technology | null | 10-20 | 10-20 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Risk & Governance Function | Less than 10% | A/B Tests for Models, Development-Staging-Production Environments | Business / Domain Practitioner | C-Suite | 35-39 | USA | Male |
| PyTorch/Lightning/Fast.ai | LLMs, Text / NLP (Non-LLM) | translation and multilinguality | Less than 3 months | Amazon Web Services | Gaps in tooling and support for model productionisation, Access to relevant data for training, Machine learning monitoring and observability | null | null | null | Custom Built In-house tool | Databricks | FastAPI/Flask Wrapper | null | Snowflake | OpenAI | Technology | 250-1,000 employees | 100-1000 | 100-1000 | Data Platform / Data Engineering Organisation | More than 90% | CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Manager | 35-39 | Czech Republic | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision, LLMs | quality assurance | Less than 6 months | Azure | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation, Machine learning monitoring and observability | Spreadsheets | null | null | null | Azure ML Studio | FastAPI/Flask Wrapper | null | Snowflake | null | Healthcare | 5,000-50,000 employees | 2-5 | 5-9 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Risk & Governance Function, AI Inventory (Keeping track of all AI usecases and models) | Between 50% and 90% | Canary Deployments, CI/CD for continuous deployment | Machine Learning Engineer | Individual Contributor (Staff+) | 40-44 | null | Male |
| Sklearn | Tabular | Fraud, Risk, Pricing | Less than a month | Amazon Web Services | Lack of specialised engineers, Building production-grade machine learning and data pipelines, Governance and domain risks | Hopsworks | Hopsworks | Hopsworks | Airflow | Hopsworks | KServe | Hopsworks | Deltalake | OpenAI | Insurance | 5,000-50,000 employees | 21-100 | 100-1000 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Less than 10% | A/B Tests for Models, CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Staff+) | 45-49 | Northern Ireland | Male |
| Sklearn | Time Series / Forecasting | Demand Forecasting, Anomaly Détection | Less than a month | Azure | Access to relevant data for training, Showcasing business impact and business value, Machine learning monitoring and observability | MLFlow | null | null | null | null | Custom Built In-house tool | Custom Built In-house tool | null | null | Energy | 50-250 employees | 100-1000 | 1000+ | Central Machine Learning Platform / Team | Less than 10% | null | Data Engineer | Individual Contributor (Staff+) | 30-34 | France | Male |
| Sklearn | Tabular, Time Series / Forecasting | Risk | Less than 6 months | Google Cloud Platform | Building production-grade machine learning and data pipelines, Machine learning security | Data Version Control (DVC) | null | null | Abinitio | null | FastAPI/Flask Wrapper | Evidently AI | GCP / BigLake | null | Financial services | 50-250 employees | null | null | null | null | null | Data Scientist | Individual Contributor (Junior to Senior) | 30-34 | null | Male |
| PyTorch/Lightning/Fast.ai | Causal Inference, LLMs, Recommender Systems, Search | Search, Recommender systems | Less than a month | Amazon Web Services | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation, Showcasing business impact and business value | MLFlow | Tecton | Custom Built In-house tool | Databricks | Custom Built In-house tool | Sagemaker | Custom Built In-house tool | Deltalake | Amazon Bedrock | Technology | 5,000-50,000 employees | 100-1000 | 100-1000 | Central Machine Learning Platform / Team, AI Inventory (Keeping track of all AI usecases and models) | Between 50% and 90% | Canary Deployments, A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 30-34 | USA | Male |
| Sklearn | LLMs, Recommender Systems, Search, Text / NLP (Non-LLM) | Search, Recommender systems | Less than a month | Amazon Web Services | Access to relevant data for training, Showcasing business impact and business value, Building production-grade machine learning and data pipelines | Spreadsheets | null | opensearch / Azure AI Search | aws step function | Amazon SageMaker | FastAPI/Flask Wrapper | null | AWS / Lakeformation | Amazon Bedrock | Insurance | 5,000-50,000 employees | 10-20 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 50% and 90% | CI/CD for continuous deployment | Data Scientist | Individual Contributor (Staff+) | 30-34 | Italy | Male |
| Pymc | Causal Inference, Tabular, Time Series / Forecasting | Demand Forecasting, Decision support | Less than 6 months | Amazon Web Services | Access to relevant data for training | MLFlow | null | null | Databricks workflow | Databricks | null | null | Deltalake | null | Retail | 50-250 employees | 21-100 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Inventory (Keeping track of all AI usecases and models) | Less than 10% | Development-Staging-Production Environments | Data Scientist | Director/VP | 40-44 | null | Male |
| PyTorch/Lightning/Fast.ai | Tabular, Time Series / Forecasting | Fraud, Risk | Less than a month | Amazon Web Services | Building production-grade machine learning and data pipelines, Governance and domain risks | Optuna | Custom Built In-house tool | null | Airflow | Custom Built In-house tool | FastAPI/Flask Wrapper | Custom Built In-house tool | Custom Built In-house tool | null | Financial services | 250-1,000 employees | 5-9 | 5-9 | Data Platform / Data Engineering Organisation, AI Risk & Governance Function | More than 90% | Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 25-29 | USA | Male |
| Sklearn | LLMs, Search, Tabular, Text / NLP (Non-LLM) | Search, Digitalisation | Less than a month | Amazon Web Services | Access to relevant data for training, Unit economics | MLFlow | Custom Built In-house tool | Elasticsearch | Airflow | Databricks | null | null | Deltalake | OpenAI | Food | 50-250 employees | 1 | 2-5 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | More than 90% | A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | Data Scientist | Individual Contributor (Junior to Senior) | 30-34 | Germany | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision, LLMs, Recommender Systems, Search, Tabular, Text / NLP (Non-LLM) | Marketing Intelligence, Recommender systems, Fraud, Risk | Less than 3 months | Azure | Lack of specialised engineers, Machine learning monitoring and observability, Machine learning security | MLFlow | null | FAISS | Databricks | Databricks | FastAPI/Flask Wrapper | Custom Built In-house tool | Deltalake | OpenAI | Technology | 10-50 employees | 10-20 | 21-100 | A Developer Productivity Team which also covers machine learning, Data Platform / Data Engineering Organisation | Less than 10% | CI/CD for continuous deployment, Development-Staging-Production Environments | Data Scientist | Individual Contributor (Staff+) | 40-44 | Netherlands | Male |
| Sklearn | LLMs, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, | Less than a month | On-prem | Access to relevant data for training, Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | null | null | null | null | null | null | null | null | Google Gemini | Technology | 10-50 employees | 2-5 | 5-9 | null | Less than 10% | null | Product Manager | C-Suite | 40-44 | India | Male |
| null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
| Sklearn | Causal Inference, Image / Computer Vision, LLMs, Recommender Systems, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Search, Recommender systems, Pricing | Less than 6 months | Amazon Web Services | Access to relevant data for training, Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | MLFlow | Databricks | Databricks | Airflow | Databricks | Custom Built In-house tool | Custom Built In-house tool | Deltalake | OpenAI | Retail | 5,000-50,000 employees | 21-100 | 100-1000 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Risk & Governance Function | Less than 10% | A/B Tests for Models, CI/CD for continuous deployment | Machine Learning Engineer | Individual Contributor (Staff+) | 40-44 | Denmark | Female |
| Sklearn | Recommender Systems, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Recommender systems, Pricing, Optimization, finance | Less than 6 months | Amazon Web Services | Access to relevant data for training, Showcasing business impact and business value, Governance and domain risks | MLFlow | Custom Built In-house tool | null | Dagster | null | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | null | Pharma | 5,000-50,000 employees | 10-20 | 10-20 | A Developer Productivity Team which also covers machine learning | Less than 10% | Development-Staging-Production Environments | Machine Learning Engineer | Manager | 30-34 | Spain | Male |
| TensorFlow | Image / Computer Vision, LLMs, Text / NLP (Non-LLM) | Marketing Intelligence, Search | Less than a month | Amazon Web Services | Lack of specialised engineers | Spreadsheets | Custom Built In-house tool | Opensearch | Custom Built In-house tool | Custom Built In-house tool | FastAPI/Flask Wrapper | null | null | Amazon Bedrock | Market Research | 50-250 employees | 10-20 | 10-20 | Central Machine Learning Platform / Team | Less than 10% | CI/CD for continuous deployment | Data Scientist | Manager | 30-34 | India | Male |
| Catboost | Recommender Systems, Tabular, Text / NLP (Non-LLM) | Recommender systems, Pricing, OCR based tasks | Less than 6 months | Google Cloud Platform | Inconsistency of training and experimentation environments, Machine learning monitoring and observability, Machine learning security | Custom Built In-house tool | FEAST | null | Kubeflow | Google Cloud Vertex AI | FastAPI/Flask Wrapper | null | GCP / BigLake | null | Automotive | 1,000-5,000 employees | 10-20 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 50% and 90% | Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Junior to Senior) | 30-34 | India | Male |
| Catboost | Causal Inference, Recommender Systems, Tabular | Recommender systems, Pricing | Less than 6 months | Google Cloud Platform | Inconsistency of training and experimentation environments, Access to relevant data for training, Building production-grade machine learning and data pipelines | GCS and Clearml | FEAST | null | kubeflow and ray | Google Cloud Vertex AI | FastAPI/Flask Wrapper | null | GCP / BigLake | Google Gemini | Automotive | 1,000-5,000 employees | 21-100 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 50% and 90% | CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Staff+) | 30-34 | India | Male |
| Sklearn | Tabular, Time Series / Forecasting | Demand Forecasting | Less than 6 months | Amazon Web Services | Lack of specialised engineers, Showcasing business impact and business value | MLFlow | Custom Built In-house tool | null | Custom Built In-house tool | Custom Built In-house tool | KServe | Evidently AI | Dremio | null | Technology | 10-50 employees | 2-5 | 0 | Central Machine Learning Platform / Team | Less than 10% | CI/CD for continuous deployment | MLOps Engineer | Individual Contributor (Junior to Senior) | 30-34 | Colombia | Male |
| Catboost | Causal Inference, LLMs, Recommender Systems, Tabular, Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Search, Recommender systems, Fraud, Pricing | Less than a month | Amazon Web Services | Access to relevant data for training, Lack of specialised engineers, Machine learning monitoring and observability | MLFlow | Tecton | Weaviate | Prefect | Databricks | Sagemaker | Custom Built In-house tool | Custom Built In-house tool | Amazon Bedrock | Food | 5,000-50,000 employees | 21-100 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Less than 10% | CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Manager | 40-44 | USA | Male |
| Sklearn | Recommender Systems, Search | Search, Recommender systems | Less than a month | Amazon Web Services | Gaps in tooling and support for model productionisation, Lack of specialised data scientists, Machine learning monitoring and observability | Spreadsheets | Custom Built In-house tool | null | Airflow | Amazon SageMaker | Sagemaker | Custom Built In-house tool | Snowflake | OpenAI | Healthcare | 50-250 employees | 5-9 | 5-9 | Data Platform / Data Engineering Organisation | More than 90% | Canary Deployments, A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | Data Engineer | Director/VP | 30-34 | USA | Male |
| XGBoost | LLMs, Recommender Systems, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Marketing Intelligence, Recommender systems, Risk, Pricing | Less than a month | Google Cloud Platform | Lack of specialised engineers, Showcasing business impact and business value, Machine learning monitoring and observability | Weights & Biases | Custom Built In-house tool | null | Airflow | Google Cloud Vertex AI | null | null | GCP / BigLake | Google Gemini | Telecommunications | 50,000+ employees | 100-1000 | 100-1000 | A Developer Productivity Team which also covers machine learning, Data Platform / Data Engineering Organisation, AI Risk & Governance Function | Less than 10% | CI/CD for continuous deployment, Development-Staging-Production Environments | Business / Domain Practitioner | Manager | 45-49 | United Kingdom | Male |
| PyTorch/Lightning/Fast.ai | LLMs | null | Less than a year | Amazon Web Services | Inconsistency of training and experimentation environments, Access to relevant data for training, Lack of specialised engineers | MLFlow | Custom Built In-house tool | Milvus | Argo Workflows | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Media & Entertainment | 5,000-50,000 employees | 2-5 | 5-9 | A Developer Productivity Team which also covers machine learning | Between 10% and 30% | Canary Deployments | Machine Learning Engineer | Individual Contributor (Staff+) | 55-59 | Japan | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision, 3D data (volumetric, mesh...) | Data analysis and enhancement | More than a year | Amazon Web Services | Access to relevant data for training, Lack of specialised engineers, Lack of specialised data scientists | ClearML | null | null | ClearML | Custom Built In-house tool | FastAPI/Flask Wrapper | null | null | null | Healthcare | 10-50 employees | 5-9 | 10-20 | Central Machine Learning Platform / Team | null | null | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 25-29 | null | Female |
| XGBoost | Causal Inference, Tabular, Time Series / Forecasting | Demand Forecasting | Less than a month | Azure | Access to relevant data for training, Showcasing business impact and business value, Machine learning monitoring and observability | Custom Built In-house tool | null | null | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | null | Custom Built In-house tool | null | Insurance | 50-250 employees | 5-9 | 10-20 | null | Less than 10% | A/B Tests for Models, Development-Staging-Production Environments | Data Scientist | Individual Contributor (Junior to Senior) | 35-39 | USA | Male |
| Sklearn | Causal Inference, LLMs, Recommender Systems, Text / NLP (Non-LLM), Time Series / Forecasting, Reinforcement Learning | Demand Forecasting, Marketing Intelligence, Search, Recommender systems, Fraud, Risk, Pricing | Less than 3 months | Azure | Gaps in tooling and support for model productionisation, Machine learning monitoring and observability | MLFlow | Custom Built In-house tool | Chroma | Custom Built In-house tool | Azure ML Studio | FastAPI/Flask Wrapper | Custom Built In-house tool | Azure DataLake | OpenAI | Consultancy | 50-250 employees | 21-100 | 100-1000 | A Developer Productivity Team which also covers machine learning | Between 10% and 30% | Canary Deployments, A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Staff+) | 35-39 | United Kingdom | Non-binary |
| Nixtla, darts, sktime, gluonts | Causal Inference, Time Series / Forecasting | Demand Forecasting | Less than a week | Azure | Machine learning monitoring and observability | Custom Built In-house tool | Custom Built In-house tool | null | Argo Workflows | Kubernetes, docker | Kubernetes | NannyML | TDengine | OpenAI | Technology | Less than 10 employees | 1000+ | 1000+ | A Developer Productivity Team which also covers machine learning | Less than 10% | null | Machine Learning Engineer | Director/VP | 50-54 | France | Male |
| PyTorch/Lightning/Fast.ai | Tabular, Text / NLP (Non-LLM) | Drug Discovery | Less than a month | Amazon Web Services | Gaps in tooling and support for model productionisation, Lack of specialised engineers, Machine learning monitoring and observability | MLFlow | Custom Built In-house tool | null | Argo Workflows | Custom Built In-house tool | KServe | null | AWS / Lakeformation | Amazon Bedrock | Healthcare | 50-250 employees | 1000+ | 1000+ | A Developer Productivity Team which also covers machine learning, Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Inventory (Keeping track of all AI usecases and models) | More than 90% | Canary Deployments, A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Director/VP | 40-44 | United Kingdom | Male |
| XGBoost | LLMs, Recommender Systems | Search, Recommender systems | Less than a month | Google Cloud Platform | Inconsistency of training and experimentation environments, Showcasing business impact and business value | Weights & Biases | Custom Built In-house tool | gcp | Airflow | Google Cloud Vertex AI | Seldon Core | Arize AI | GCP / BigLake | fireworks | Technology | 50-250 employees | 2-5 | 5-9 | Central Machine Learning Platform / Team | Between 50% and 90% | A/B Tests for Models | Machine Learning Engineer | Individual Contributor (Staff+) | 35-39 | United Kingdom | Male |
| Sklearn | LLMs, Recommender Systems, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Recommender systems | Less than 3 months | Google Cloud Platform | Inconsistency of training and experimentation environments, Machine learning monitoring and observability | Custom Built In-house tool | FEAST | Pgvector | Airflow | Custom Built In-house tool | FastAPI/Flask Wrapper | null | GCP / BigLake | Azure AI | Saas | 50-250 employees | 10-20 | 21-100 | null | Between 50% and 90% | CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 35-39 | Spain | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision, LLMs, Text / NLP (Non-LLM) | Fraud, Insurance | Less than 3 months | Amazon Web Services | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation | Weights & Biases | Custom Built In-house tool | null | Meta flow | Amazon SageMaker | Sagemaker | Custom Built In-house tool | AWS / Lakeformation | OpenAI | Insurance | 50-250 employees | 100-1000 | 1000+ | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Inventory (Keeping track of all AI usecases and models) | Between 50% and 90% | Canary Deployments, A/B Tests for Models | Machine Learning Engineer | Director/VP | 35-39 | United Kingdom | Male |
| PyTorch/Lightning/Fast.ai | LLMs, Text / NLP (Non-LLM) | Classification | Less than 3 months | Amazon Web Services | Inconsistency of training and experimentation environments, Showcasing business impact and business value, Machine learning monitoring and observability | MLFlow | null | Milvus | null | Databricks | FastAPI/Flask Wrapper | null | Deltalake | Amazon Bedrock | Technology | 5,000-50,000 employees | null | null | null | null | null | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 25-29 | USA | Male |
| TensorFlow | Image / Computer Vision, LLMs, Recommender Systems, Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Recommender systems, Fraud | Less than 3 months | Google Cloud Platform | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation, Building production-grade machine learning and data pipelines | MLFlow | Hopsworks | Pinecone | Airflow | Google Cloud Vertex AI | KServe | Evidently AI | GCP / BigLake | Google Gemini | Financial services | 50,000+ employees | 21-100 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 50% and 90% | A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Staff+) | 30-34 | Netherlands | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision, LLMs, Text / NLP (Non-LLM) | Recommender systems, Fraud | Less than 3 months | Amazon Web Services | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation, Access to relevant data for training | Weights & Biases | Custom Built In-house tool | null | Custom Built In-house tool | Amazon SageMaker | Sagemaker | Custom Built In-house tool | AWS / Lakeformation | Amazon Bedrock | Insurance | 50-250 employees | 100-1000 | 1000+ | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 50% and 90% | Canary Deployments, A/B Tests for Models | Machine Learning Engineer | Director/VP | 35-39 | United Kingdom | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision, LLMs, Recommender Systems, Search | Search, Recommender systems | Less than 3 months | Azure | Gaps in tooling and support for model productionisation, Building production-grade machine learning and data pipelines, Governance and domain risks | MLFlow | null | Elasticsearch | Azure ML pipelines | Azure ML Studio | Azure Managed Online Endpoints | Custom Built In-house tool | null | Azure AI | Public Sector | 1,000-5,000 employees | 10-20 | 21-100 | null | Between 50% and 90% | Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Staff+) | 30-34 | null | null |
| gpflow | Causal Inference, Recommender Systems, Tabular, Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, process optimization | More than a year | Amazon Web Services | Gaps in tooling and support for model productionisation, Access to relevant data for training, Building production-grade machine learning and data pipelines | MLFlow | Custom Built In-house tool | None | Airflow | Custom Built In-house tool | FastAPI/Flask Wrapper | null | AWS / Lakeformation | None | Automotive | 5,000-50,000 employees | 21-100 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 50% and 90% | null | MLOps Engineer | Individual Contributor (Junior to Senior) | 35-39 | Luxembourg | Male |
| XGBoost | Tabular, Time Series / Forecasting | Demand Forecasting, Cost, stock, project duration forecasting | Less than a month | Amazon Web Services | Lack of specialised engineers, Lack of specialised data scientists | MLFlow | Custom Built In-house tool | None | Airflow | Custom Built In-house tool | FastAPI/Flask Wrapper | Evidently AI | AWS / Lakeformation | Custom Built In-house tool | Consultancy | 250-1,000 employees | 5-9 | 10-20 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Less than 10% | Canary Deployments, Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Junior to Senior) | 30-34 | Spain | Male |
| Sklearn | Tabular | Fraud | Less than 3 months | Amazon Web Services | Building production-grade machine learning and data pipelines, Machine learning monitoring and observability, Inconsistency of training and production environments | Custom Built In-house tool | Custom Built In-house tool | null | Airflow | Custom Built In-house tool | Sagemaker | null | Snowflake | null | Financial services | 50-250 employees | 10-20 | 2-5 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | More than 90% | null | Data Scientist | Individual Contributor (Staff+) | 25-29 | Germany | Male |
| Non | Image / Computer Vision, LLMs | Digitalisation | Less than 3 months | Amazon Web Services | Building production-grade machine learning and data pipelines | null | null | PostgreSQL | null | null | AWS lambda | null | null | Google Gemini | Technology | Less than 10 employees | 1 | 1 | A Developer Productivity Team which also covers machine learning | Between 10% and 30% | null | Software Engineer | Individual Contributor (Staff+) | 25-29 | India | Male |
| Sklearn | Causal Inference, Recommender Systems, Tabular, Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Recommender systems, Fraud, Risk, Pricing | Less than 3 months | Amazon Web Services | Gaps in tooling and support for model productionisation, Access to relevant data for training, Machine learning monitoring and observability | MLFlow | null | null | Airflow | Custom Built In-house tool | FastAPI/Flask Wrapper | null | null | null | Technology | 5,000-50,000 employees | 100-1000 | 1000+ | A Developer Productivity Team which also covers machine learning, Data Platform / Data Engineering Organisation, AI Risk & Governance Function, AI Inventory (Keeping track of all AI usecases and models) | Less than 10% | CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Staff+) | 35-39 | null | Male |
| PyTorch/Lightning/Fast.ai | LLMs, Text / NLP (Non-LLM) | Recommender systems | Less than 3 months | Amazon Web Services | Access to relevant data for training, Building production-grade machine learning and data pipelines, Machine learning security | null | null | null | null | null | null | null | null | null | Healthcare | 10-50 employees | 0 | 1 | A Developer Productivity Team which also covers machine learning | Less than 10% | null | Software Engineer | Individual Contributor (Junior to Senior) | 22-24 | india | Male |
| null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
| null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
| PyTorch/Lightning/Fast.ai | Text / NLP (Non-LLM) | Recommender systems | Less than 6 months | Amazon Web Services | Gaps in tooling and support for model productionisation, Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | Sagemaker | Fennel | Pinecone | Airflow | Amazon SageMaker | Sagemaker | null | AWS / Lakeformation | Anthropic | Hospitality | Less than 10 employees | 2-5 | 5-9 | A Developer Productivity Team which also covers machine learning, Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 10% and 30% | CI/CD for continuous deployment | MLOps Engineer | Individual Contributor (Staff+) | 45-49 | Netherlands | Male |
| Sklearn | Tabular | Risk | Less than 6 months | Amazon Web Services | Lack of specialised data scientists, Showcasing business impact and business value | MLFlow | Custom Built In-house tool | No vector database | Airflow | Custom Built In-house tool | FastAPI/Flask Wrapper | Model metrics monitoring on Metabase | Snowflake | Amazon Bedrock | Financial services | 10-50 employees | 2-5 | 1 | Central Machine Learning Platform / Team | More than 90% | A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment | Data Scientist | Director/VP | 30-34 | France | Male |
| Sklearn | Recommender Systems, Time Series / Forecasting | Demand Forecasting, Recommender systems | More than a year | Azure | Inconsistency of training and experimentation environments, Showcasing business impact and business value, Building production-grade machine learning and data pipelines | MLFlow | Custom Built In-house tool | Azure Cognitive Search | Databricks Workflows | Databricks | FastAPI/Flask Wrapper | null | Deltalake | Azure AI | Retail | 50,000+ employees | 5-9 | 10-20 | Data Platform / Data Engineering Organisation | Between 10% and 30% | Canary Deployments, A/B Tests for Models, CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 25-29 | Netherlands | Male |
| Sklearn | Causal Inference, LLMs, Recommender Systems, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Recommender systems, Risk | Less than 6 months | Databricks | Gaps in tooling and support for model productionisation, Machine learning monitoring and observability, Governance and domain risks | MLFlow | Custom Built In-house tool | ElasticSearch | Custom Built In-house tool | Databricks | FastAPI/Flask Wrapper | null | Snowflake | Azure AI | Human Resources | 50,000+ employees | 10-20 | 5-9 | Central Machine Learning Platform / Team | Less than 10% | CI/CD for continuous deployment | Data Scientist | Manager | 30-34 | USA | Male |
| Langchain | LLMs | Agents | Less than 3 months | Amazon Web Services | Inconsistency of training and experimentation environments, Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | Weights & Biases | None | Pgvector | Airflow | Weights and biases | FastAPI/Flask Wrapper | Arize AI | Snowflake | OpenAI | Technology | 50-250 employees | 2-5 | 5-9 | Data Platform / Data Engineering Organisation | Between 50% and 90% | None | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 25-29 | USA | Male |
| TensorFlow | Time Series / Forecasting | Meteorology | Less than a year | Amazon Web Services | Lack of specialised engineers, Machine learning monitoring and observability | MLFlow | Custom Built In-house tool | null | snakemake | null | null | null | Custom Built In-house tool | null | Technology | 250-1,000 employees | null | null | Central Machine Learning Platform / Team, AI Inventory (Keeping track of all AI usecases and models) | null | null | Data Scientist | Individual Contributor (Junior to Senior) | 25-29 | Switzerland | Male |
| PyTorch/Lightning/Fast.ai | Time Series / Forecasting | Risk | Less than 3 months | On-prem | Inconsistency of training and experimentation environments, Access to relevant data for training, Machine learning monitoring and observability | MLFlow | Custom Built In-house tool | Custom Built In-house tool | Airflow | Custom Built In-house tool | FastAPI/Flask Wrapper | Neptune AI | Custom Built In-house tool | null | Enviroment | 250-1,000 employees | 10-20 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | More than 90% | CI/CD for continuous deployment | Data Scientist | Individual Contributor (Junior to Senior) | 30-34 | Switzerland | Male |
| Sklearn | Image / Computer Vision, LLMs, Recommender Systems, Search, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Marketing Intelligence, Search, Fraud | Less than 3 months | Azure | Lack of specialised engineers, Lack of specialised data scientists, Machine learning monitoring and observability | MLFlow | None | ChromaDB | Custom Built In-house tool | Custom Built In-house tool | FastAPI/Flask Wrapper | Evidently AI | Deltalake | OpenAI | Consulting | 1,000-5,000 employees | 21-100 | 21-100 | A Developer Productivity Team which also covers machine learning, Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Risk & Governance Function | Less than 10% | A/B Tests for Models | Machine Learning Engineer | Director/VP | 35-39 | France | Male |
| Catboost | Image / Computer Vision, LLMs, Tabular, Time Series / Forecasting | Defect detection | Less than a month | Amazon Web Services | Access to relevant data for training, Lack of specialised data scientists, Building production-grade machine learning and data pipelines | MLFlow | Databricks | Databricks | Airflow | Databricks | Databricks | Custom Built In-house tool | Deltalake | OpenAI | Manufacturing | 1,000-5,000 employees | 5-9 | 10-20 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | More than 90% | A/B Tests for Models, CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Staff+) | 30-34 | null | null |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision, LLMs, Search | Search, Augmentation | Less than 3 months | Amazon Web Services | Gaps in tooling and support for model productionisation, Lack of specialised engineers, Building production-grade machine learning and data pipelines | Weights & Biases | null | Elasticsearch / Opensearch | Amazon Glue | Custom Built In-house tool | FastAPI/Flask Wrapper | null | null | OpenAI | Technology | Less than 10 employees | 1 | 2-5 | null | null | CI/CD for continuous deployment | Machine Learning Engineer | Individual Contributor (Staff+) | 30-34 | United Kingdom | Female |
| null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision, Recommender Systems, Tabular | Recommender systems | Less than 3 months | On-prem | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation | ClearML | FEAST | Milvus | kubeflow pipelines | kubeflow | KServe | Custom Built In-house tool | Snowflake | Custom Built In-house tool | Media & Entertainment | 250-1,000 employees | 10-20 | 10-20 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 50% and 90% | A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment | Machine Learning Engineer | Individual Contributor (Junior to Senior) | null | null | null |
| XGBoost | Image / Computer Vision, Tabular, Text / NLP (Non-LLM) | Fraud, Risk | Less than 3 months | Google Cloud Platform | Access to relevant data for training, Lack of specialised data scientists, Showcasing business impact and business value | MLFlow | null | null | Airflow | Google Cloud Vertex AI | Spring Boot | Custom Built In-house tool | Custom Built In-house tool | null | Retail | 1,000-5,000 employees | 21-100 | 10-20 | A Developer Productivity Team which also covers machine learning | Less than 10% | Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 30-34 | Netherlands | Male |
| null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
| XGBoost | Image / Computer Vision, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Fraud, Risk, Customer Satisfaction | More than a year | Azure | Gaps in tooling and support for model productionisation, Lack of specialised engineers, Governance and domain risks | MLFlow | Databricks | null | Databricks | Databricks | FastAPI/Flask Wrapper | Custom Built In-house tool | Azure DataLake | OpenAI | Government | 5,000-50,000 employees | 5-9 | 10-20 | Data Platform / Data Engineering Organisation | Between 10% and 30% | CI/CD for continuous deployment, Development-Staging-Production Environments | Data Scientist | Individual Contributor (Staff+) | 35-39 | UK | Male |
| TensorFlow | Recommender Systems, Search | Search, Recommender systems | Less than 6 months | GCP and AWS equally | Gaps in tooling and support for model productionisation, Showcasing business impact and business value, Big organisational migrations (of Datasets, MLP and Data platforms and mroe) | Google MetaDataStore | Custom Built In-house tool | Custom Built In-house tool | Airflow | Google Cloud Vertex AI | Sagemaker | Evidently AI | GCP / BigLake | null | Food Delivery | 5,000-50,000 employees | 100-1000 | 100-1000 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Risk & Governance Function | Less than 10% | Canary Deployments, A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 30-34 | Netherlands | Male |
| langchain | Image / Computer Vision, LLMs | Search, Recommender systems, Chatbot | Less than 3 months | Amazon Web Services | Showcasing business impact and business value, Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | MLFlow | Custom Built In-house tool | MongoDB | Airflow | Databricks | Databricks | Custom Built In-house tool | Deltalake | Azure AI | Retail | 5,000-50,000 employees | 21-100 | 21-100 | A Developer Productivity Team which also covers machine learning, Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Risk & Governance Function | Between 10% and 30% | A/B Tests for Models, CI/CD for continuous deployment | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 25-29 | Spain | Male |
| Sklearn | LLMs | Search | Less than a month | Azure | Gaps in tooling and support for model productionisation, Lack of specialised engineers, Machine learning monitoring and observability | MLFlow | Custom Built In-house tool | Pgvector | null | Custom Built In-house tool | null | null | Azure DataLake | Azure AI | Transportation & warehousing | 50,000+ employees | 10-20 | 5-9 | A Developer Productivity Team which also covers machine learning, Data Platform / Data Engineering Organisation, AI Risk & Governance Function | Between 10% and 30% | Progressive Rollouts | MLOps Engineer | Individual Contributor (Junior to Senior) | 30-34 | portugal | Male |
| XGBoost | Causal Inference, Tabular, Time Series / Forecasting | Fraud, Risk | Less than a year | null | Access to relevant data for training, Showcasing business impact and business value, Machine learning monitoring and observability | Custom Built In-house tool | null | null | null | Custom Built In-house tool | FastAPI/Flask Wrapper | null | Custom Built In-house tool | null | Media & Entertainment | 250-1,000 employees | 5-9 | 10-20 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 10% and 30% | A/B Tests for Models | Data Scientist | Manager | null | null | Male |
| Sklearn | LLMs, Recommender Systems, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Recommender systems, Pricing | Less than 6 months | Databricks | Access to relevant data for training, Machine learning monitoring and observability | MLFlow | Databricks | null | Airflow | Databricks | Databricks | Evidently AI | Deltalake | OpenAI | Media & Entertainment | 1,000-5,000 employees | 10-20 | 21-100 | Data Platform / Data Engineering Organisation, AI Inventory (Keeping track of all AI usecases and models) | Between 10% and 30% | Canary Deployments, A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment | Machine Learning Engineer | Manager | 35-39 | Canada | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision | null | Less than a month | Amazon Web Services | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation, Building production-grade machine learning and data pipelines | null | null | null | null | Amazon SageMaker | Sagemaker | null | null | null | Healthcare | Less than 10 employees | null | null | null | null | null | null | null | null | null | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision, LLMs | Search, Information extraction | Less than 6 months | Amazon Web Services | Gaps in tooling and support for model productionisation, Machine learning monitoring and observability, Governance and domain risks | Custom Built In-house tool | null | null | Prefect | Custom Built In-house tool | null | null | null | null | Consultancy | 50-250 employees | 2-5 | 2-5 | Central Machine Learning Platform / Team | Less than 10% | CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 30-34 | germany | Male |
| PyTorch/Lightning/Fast.ai | Recommender Systems | Recommender systems | Less than 3 months | Amazon Web Services | Gaps in tooling and support for model productionisation, Lack of specialised engineers | Weights & Biases | null | Elasticsearch | Custom Built In-house tool | Custom Built In-house tool | Sagemaker | null | GCP / BigLake | OpenAI | Media & Entertainment | 1,000-5,000 employees | 2-5 | 2-5 | A Developer Productivity Team which also covers machine learning | Between 50% and 90% | A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 30-34 | Japan | Male |
| Sklearn | LLMs, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Fraud, Risk, Pricing | Less than 3 months | Amazon Web Services | Inconsistency of training and experimentation environments, Machine learning monitoring and observability, Governance and domain risks | Weights & Biases | Custom Built In-house tool | Pinecone | Amazon Glue | Amazon SageMaker | FastAPI/Flask Wrapper | Neptune AI | Snowflake | Amazon Bedrock | Technology | 50,000+ employees | 100-1000 | 1000+ | A Developer Productivity Team which also covers machine learning, Data Platform / Data Engineering Organisation, AI Risk & Governance Function | Between 50% and 90% | Canary Deployments, CI/CD for continuous deployment, Development-Staging-Production Environments | Data Analyst | Individual Contributor (Junior to Senior) | 40-44 | India | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision | robotics | Less than 3 months | Google Cloud Platform | Access to relevant data for training, Lack of specialised data scientists, Machine learning monitoring and observability | ClearML | null | null | Airflow | Custom Built In-house tool | tensorrt | null | Custom Built In-house tool | Anthropic | Transportation & warehousing | Less than 10 employees | 2-5 | 2-5 | Central Machine Learning Platform / Team | More than 90% | CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Staff+) | 30-34 | netherlands | Male |
| Sklearn | LLMs, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Pricing | Less than a year | IBM | Inconsistency of training and experimentation environments, Lack of specialised engineers, Machine learning monitoring and observability | MLFlow | Custom Built In-house tool | Weaviate | Custom Built In-house tool | Domino | FastAPI/Flask Wrapper | Custom Built In-house tool | IBM | Azure AI | Financial services | 5,000-50,000 employees | 10-20 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 10% and 30% | Canary Deployments, A/B Tests for Models, CI/CD for continuous deployment | Data Scientist | Individual Contributor (Staff+) | 30-34 | FRANCE | Female |
| XGBoost | LLMs, Recommender Systems, Tabular | Search, Recommender systems, content generation | Less than 6 months | Amazon Web Services | Gaps in tooling and support for model productionisation, Access to relevant data for training, Showcasing business impact and business value | Spreadsheets | Custom Built In-house tool | null | Airflow | Custom Built In-house tool | Custom Built In-house tool | null | Snowflake | OpenAI | Media & Entertainment | 50-250 employees | 2-5 | 5-9 | A Developer Productivity Team which also covers machine learning, Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 10% and 30% | A/B Tests for Models | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 25-29 | Finland | Male |
| XGBoost | Tabular | Fraud, Risk, Pricing | Less than 6 months | Azure | Lack of specialised data scientists, Machine learning monitoring and observability, Governance and domain risks | Custom Built In-house tool | null | null | Argo Workflows | Custom Built In-house tool | FastAPI/Flask Wrapper | null | null | null | Financial services | 10-50 employees | 2-5 | 5-9 | Central Machine Learning Platform / Team, AI Inventory (Keeping track of all AI usecases and models) | Less than 10% | Development-Staging-Production Environments | Machine Learning Engineer | Director/VP | 45-49 | Israel | Male |
| Sklearn | Tabular | null | Less than a month | Azure | Access to relevant data for training, Lack of specialised engineers, Machine learning monitoring and observability | MLFlow | null | null | Prefect | Azure ML Studio | Custom Built In-house tool | null | Azure DataLake | null | Technology | 1,000-5,000 employees | 21-100 | 100-1000 | null | Less than 10% | CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Junior to Senior) | 30-34 | United Kingdom | Male |
| PyTorch/Lightning/Fast.ai | Recommender Systems | Recommender systems | Less than a month | Google Cloud Platform | Gaps in tooling and support for model productionisation | MLFlow | Custom Built In-house tool | Qdrant | databricks | databricks | triton | Arize AI | Deltalake | null | Retail | 1,000-5,000 employees | 21-100 | 100-1000 | Central Machine Learning Platform / Team | Between 50% and 90% | A/B Tests for Models | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 25-29 | india | Male |
| PyTorch/Lightning/Fast.ai | Text / NLP (Non-LLM) | Fraud | Less than 3 months | Google Cloud Platform | Gaps in tooling and support for model productionisation, Access to relevant data for training, Building production-grade machine learning and data pipelines | MLFlow | Hopsworks | Pinecone | Airflow | Amazon SageMaker | FastAPI/Flask Wrapper | Neptune AI | GCP / BigLake | Anthropic | Financial services | 10-50 employees | 2-5 | 5-9 | A Developer Productivity Team which also covers machine learning | Between 10% and 30% | Progressive Rollouts | Machine Learning Engineer | Manager | 30-34 | UAE | Male |
| TensorFlow | LLMs, Tabular, Text / NLP (Non-LLM) | Demand Forecasting, Pricing, Document entity extraction | Less than a month | Amazon Web Services | Inconsistency of training and experimentation environments, Showcasing business impact and business value, Model consistency (re-training / catastrophic forgetting) | Weights & Biases | Custom Built In-house tool | LanceDB / Turbopuffer | Metaflow | Custom Built In-house tool | FastAPI/Flask Wrapper | Evidently AI | Snowflake | OpenAI | Financial services | 1,000-5,000 employees | 1000+ | 100-1000 | A Developer Productivity Team which also covers machine learning, Data Platform / Data Engineering Organisation, AI Inventory (Keeping track of all AI usecases and models) | Less than 10% | Progressive Rollouts | Machine Learning Engineer | Individual Contributor (Staff+) | 30-34 | United States | Male |
| PyTorch/Lightning/Fast.ai | LLMs | . | Less than a month | On-prem | Gaps in tooling and support for model productionisation, Access to relevant data for training, Showcasing business impact and business value | Weights & Biases | null | Pinecone | Argo Workflows | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | Media & Entertainment | 250-1,000 employees | 21-100 | 0 | AI Inventory (Keeping track of all AI usecases and models) | Between 10% and 30% | Canary Deployments | MLOps Engineer | Manager | 55-59 | Japan | Male |
| PyTorch/Lightning/Fast.ai | LLMs | Recommender systems | Less than 6 months | Google Cloud Platform | Showcasing business impact and business value, Building production-grade machine learning and data pipelines, Governance and domain risks | Custom Built In-house tool | null | null | null | null | Seldon Core | null | null | null | Telecommunications | 10-50 employees | null | null | Central Machine Learning Platform / Team | Between 10% and 30% | null | MLOps Engineer | Manager | null | null | Male |
| Catboost | Causal Inference, LLMs, Tabular | Recommender systems, Risk, Text Summarization | Less than a month | Amazon Web Services | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation, Lack of specialised data scientists | MLFlow | DynamoDB | NA | Airflow | Local | FastAPI/Flask Wrapper | NannyML | AWS / Lakeformation | OpenAI | Healthcare | 250-1,000 employees | 2-5 | 5-9 | Data Platform / Data Engineering Organisation | Between 50% and 90% | Canary Deployments, Progressive Rollouts, CI/CD for continuous deployment | Data Scientist | Individual Contributor (Junior to Senior) | 30-34 | Brazil | Male |
| PyTorch/Lightning/Fast.ai | Causal Inference, Image / Computer Vision | Av | Less than 3 months | On-prem | Lack of specialised engineers, Lack of specialised data scientists, Building production-grade machine learning and data pipelines | Comet | Custom Built In-house tool | null | Argo Workflows | Custom Built In-house tool | null | null | AWS / Lakeformation | null | Transportation & warehousing | 1,000-5,000 employees | 100-1000 | 1000+ | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | null | Progressive Rollouts | Product Manager | Director/VP | 45-49 | null | Male |
| Sklearn | Tabular, Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Sustainability | null | Azure | Showcasing business impact and business value, Governance and domain risks | Azure ML Service, Sagemaker | null | null | null | Azure ML Studio | null | Evidently AI | null | OpenAI | Consulting | 10-50 employees | null | null | A Developer Productivity Team which also covers machine learning, Data Platform / Data Engineering Organisation | Less than 10% | null | Data Scientist | Individual Contributor (Junior to Senior) | null | null | null |
| TensorFlow | Recommender Systems, Search | Search, Recommender systems | Less than a month | AWS & GCP | Gaps in tooling and support for model productionisation, Machine learning monitoring and observability | Custom Built In-house tool | DynamoDB | Custom Built In-house tool | Airflow | Google Cloud Vertex AI | Sagemaker | null | GCP / BigLake | Amazon Bedrock | Food | 250-1,000 employees | 2-5 | 5-9 | Central Machine Learning Platform / Team | Less than 10% | Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Staff+) | null | null | Male |
| XGBoost | Causal Inference, Recommender Systems, Text / NLP (Non-LLM) | Recommender systems | Less than a year | Azure | Access to relevant data for training, Showcasing business impact and business value, Governance and domain risks | MLFlow | FEAST | null | Airflow | Custom Built In-house tool | FastAPI/Flask Wrapper | Evidently AI | null | null | Healthcare | 1,000-5,000 employees | 2-5 | 2-5 | Central Machine Learning Platform / Team | More than 90% | CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 35-39 | Netherlands | Female |
| lightgbm | LLMs, Recommender Systems, Time Series / Forecasting | Demand Forecasting, Recommender systems, Pricing | Less than 6 months | Google Cloud Platform | Lack of specialised data scientists, Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | null | Custom Built In-house tool | Weaviate | Airflow | Google Cloud Vertex AI | FastAPI/Flask Wrapper | Custom Built In-house tool | GCP / BigLake | OpenAI | Retail | 50-250 employees | 10-20 | 21-100 | A Developer Productivity Team which also covers machine learning, Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Less than 10% | A/B Tests for Models, CI/CD for continuous deployment, Development-Staging-Production Environments | Data Scientist | Individual Contributor (Staff+) | 30-34 | Germany | Male |
| PyTorch/Lightning/Fast.ai | LLMs, Recommender Systems, Text / NLP (Non-LLM) | Recommender systems, Fraud, Risk | Less than 6 months | Azure | Inconsistency of training and experimentation environments, Access to relevant data for training, Machine learning monitoring and observability | MLFlow | null | Azure ai search | Custom Built In-house tool | Databricks | Databricks | Custom Built In-house tool | Azure DataLake | Azure AI | Healthcare | 5,000-50,000 employees | 10-20 | 21-100 | A Developer Productivity Team which also covers machine learning, Data Platform / Data Engineering Organisation, AI Risk & Governance Function | Less than 10% | A/B Tests for Models, CI/CD for continuous deployment | null | Director/VP | 45-49 | null | Male |
| Sklearn | Graphs | Fraud | Less than 6 months | Amazon Web Services | Inconsistency of training and experimentation environments, Building production-grade machine learning and data pipelines | Custom Built In-house tool | Custom Built In-house tool | null | Argo Workflows | Databricks | FastAPI/Flask Wrapper | Custom Built In-house tool | AWS / Lakeformation | OpenAI | Financial services | 5,000-50,000 employees | 100-1000 | 100-1000 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Risk & Governance Function, AI Inventory (Keeping track of all AI usecases and models) | More than 90% | A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment | Data Scientist | Manager | 35-39 | Brazil | Male |
| PyTorch/Lightning/Fast.ai | LLMs, Text / NLP (Non-LLM) | Search | null | null | null | Weights & Biases | null | null | null | Custom Built In-house tool | null | null | null | null | Technology | 50-250 employees | null | null | Central Machine Learning Platform / Team | null | null | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 22-24 | The Netherlands | Male |
| Sklearn | Causal Inference, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting | Less than a month | Azure | Lack of specialised engineers, Building production-grade machine learning and data pipelines | null | null | null | Airflow | null | null | null | MSSQL | Azure AI | Healthcare | 5,000-50,000 employees | 2-5 | 5-9 | Data Platform / Data Engineering Organisation, AI Risk & Governance Function, AI Inventory (Keeping track of all AI usecases and models) | Less than 10% | null | Data Scientist | Individual Contributor (Junior to Senior) | 30-34 | US | Male |
| Catboost | Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Recommender systems, Fraud, Risk | Less than a week | Google Cloud Platform | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation, Showcasing business impact and business value | MLFlow | FEAST | null | Argo Workflows | Custom Built In-house tool | FastAPI/Flask Wrapper | Custom Built In-house tool | Custom Built In-house tool | null | Financial services | 1,000-5,000 employees | 100-1000 | 1000+ | Data Platform / Data Engineering Organisation | null | null | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 30-34 | null | Male |
| XGBoost | Time Series / Forecasting | Fraud, Risk | Less than a month | Amazon Web Services | Building production-grade machine learning and data pipelines, Governance and domain risks | MLFlow | null | null | Airflow | Metaflow | FastAPI/Flask Wrapper | Arize AI | Deltalake | null | Financial services | 1,000-5,000 employees | 21-100 | 100-1000 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Risk & Governance Function | More than 90% | Canary Deployments, A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Junior to Senior) | 30-34 | United States | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Generating synthetic data | Less than 3 months | Google Cloud Platform | Gaps in tooling and support for model productionisation, Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | Spreadsheets | null | null | Github actions | Google Cloud Vertex AI | FastAPI/Flask Wrapper | null | GCP / BigLake | null | Energy | 10-50 employees | 2-5 | 5-9 | Central Machine Learning Platform / Team | More than 90% | CI/CD for continuous deployment | Data Scientist | Individual Contributor (Junior to Senior) | 30-34 | United Kingdom | Male |
| XGBoost | Causal Inference, LLMs, Recommender Systems, Search, Tabular, Text / NLP (Non-LLM), Time Series / Forecasting | Demand Forecasting, Marketing Intelligence, Recommender systems, Fraud, Risk, Pricing | Less than a week | Azure | Inconsistency of training and experimentation environments, Lack of specialised engineers, Machine learning monitoring and observability | omega-ml | Custom Built In-house tool | Postgresql | Celery | Omega-ml | Omega-ml | Omega-ml | Custom Built In-house tool | Omega-ml | Financial services | 250-1,000 employees | 5-9 | 10-20 | Central Machine Learning Platform / Team, AI Risk & Governance Function, AI Inventory (Keeping track of all AI usecases and models) | More than 90% | Canary Deployments, A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Staff+) | 50-54 | Switzerland | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision, Search, Text / NLP (Non-LLM) | Search, biotech use cases | Less than 6 months | Google Cloud Platform | Gaps in tooling and support for model productionisation, Machine learning monitoring and observability, Governance and domain risks | Weights & Biases | null | null | Airflow | Custom Built In-house tool | Anyscale/Ray | null | GCP / BigLake | Google Gemini | Pharmaceuticals | 250-1,000 employees | 21-100 | 100-1000 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Less than 10% | Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Junior to Senior) | Prefer not to share | null | Female |
| PyTorch/Lightning/Fast.ai | Causal Inference, Image / Computer Vision, LLMs, Tabular, Text / NLP (Non-LLM) | Marketing Intelligence, Fraud, ICR/OCR | Less than 3 months | Azure | Inconsistency of training and experimentation environments, Gaps in tooling and support for model productionisation, Building production-grade machine learning and data pipelines | Custom Built In-house tool | null | null | Custom Built In-house tool | Azure ML Studio | FastAPI/Flask Wrapper | null | null | OpenAI | Insurance | 1,000-5,000 employees | 21-100 | 21-100 | Central Machine Learning Platform / Team, AI Inventory (Keeping track of all AI usecases and models) | Between 10% and 30% | CI/CD for continuous deployment | Data Scientist | Manager | 40-44 | Portugal | Male |
| PyTorch/Lightning/Fast.ai | LLMs, Text / NLP (Non-LLM) | Education | Less than 6 months | On-prem | Access to relevant data for training, Machine learning monitoring and observability | Weights & Biases | null | null | null | Google Cloud Vertex AI | null | null | null | Google Gemini | Technology | 1,000-5,000 employees | 10-20 | null | Central Machine Learning Platform / Team | Less than 10% | null | Software Engineer | Individual Contributor (Staff+) | 45-49 | Spain | Female |
| PyTorch/Lightning/Fast.ai | Text / NLP (Non-LLM) | null | More than a year | Amazon Web Services | Lack of specialised engineers, Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | Custom Built In-house tool | null | Milvus | Custom Built In-house tool | Amazon SageMaker | Nvidia Triton Inference Server | null | null | null | Retail | 50-250 employees | 2-5 | 5-9 | null | More than 90% | CI/CD for continuous deployment | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 35-39 | Turkey | Male |
| PyTorch/Lightning/Fast.ai | LLMs, Search, Text / NLP (Non-LLM) | Search, Fraud, Risk | Less than 3 months | Amazon Web Services | Machine learning monitoring and observability, Governance and domain risks, client requirements | MLFlow | null | opensearch, elasticsearch | Airflow | Databricks | FastAPI/Flask Wrapper | null | Deltalake | Azure AI | Financial services | 5,000-50,000 employees | 100-1000 | 100-1000 | Data Platform / Data Engineering Organisation, AI Risk & Governance Function | Between 10% and 30% | null | Machine Learning Engineer | Director/VP | 45-49 | USA | Male |
| PyTorch/Lightning/Fast.ai | Image / Computer Vision, Search | Search | Less than 6 months | Google Cloud Platform | Gaps in tooling and support for model productionisation, Lack of specialised engineers, Lack of specialised data scientists | MLFlow | null | null | kubeflow | Google Cloud Vertex AI | TorchServe | null | null | OpenAI | null | 250-1,000 employees | 10-20 | 5-9 | Central Machine Learning Platform / Team | Between 50% and 90% | CI/CD for continuous deployment, Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Junior to Senior) | 25-29 | null | Male |
| PyTorch/Lightning/Fast.ai | Tabular, Time Series / Forecasting | Demand Forecasting, Pricing | Less than 6 months | Amazon Web Services | Gaps in tooling and support for model productionisation, Lack of specialised engineers, Machine learning monitoring and observability | MLFlow | null | Pinecone | Airflow | Databricks | FastAPI/Flask Wrapper | Evidently AI | Deltalake | Anthropic | Retail | 1,000-5,000 employees | 10-20 | 2-5 | A Developer Productivity Team which also covers machine learning, Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Inventory (Keeping track of all AI usecases and models) | Between 10% and 30% | A/B Tests for Models, CI/CD for continuous deployment, Development-Staging-Production Environments | Data Scientist | Individual Contributor (Staff+) | 25-29 | France | Male |
| Sklearn | Text / NLP (Non-LLM) | Demand Forecasting, Marketing Intelligence, Recommender systems, Fraud, Risk, Pricing, speech recognition, document (pii) anonymisation | Less than a year | Azure | Showcasing business impact and business value, lack of vision (what happens after productionising), reservedness from stakeholders | MLFlow | none | azure's offering in ai search | Celery | Azure ML Studio | Azure ML | Evidently AI | Azure DataLake | Azure AI | Insurance | 5,000-50,000 employees | 2-5 | 10-20 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Risk & Governance Function, AI Inventory (Keeping track of all AI usecases and models) | Between 10% and 30% | Canary Deployments, A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment | MLOps Engineer | Individual Contributor (Staff+) | 35-39 | nl | Male |
| PyTorch/Lightning/Fast.ai | LLMs, Search, Text / NLP (Non-LLM) | Search | Less than 6 months | Amazon Web Services | Gaps in tooling and support for model productionisation, Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | Spreadsheets | null | OpenSearch | Custom Built In-house tool | Amazon SageMaker | FastAPI/Flask Wrapper | Custom Built In-house tool | null | Amazon Bedrock | Financial services | 1,000-5,000 employees | 10-20 | 10-20 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Risk & Governance Function | Between 10% and 30% | A/B Tests for Models, CI/CD for continuous deployment | Machine Learning Engineer | Individual Contributor (Junior to Senior) | 50-54 | USA | Male |
| LightGBM | Tabular | Fraud | Less than 6 months | Amazon Web Services | Gaps in tooling and support for model productionisation, Access to relevant data for training, Building production-grade machine learning and data pipelines | null | null | null | Airflow | null | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | null | Financial services | 250-1,000 employees | 100-1000 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation, AI Risk & Governance Function | More than 90% | A/B Tests for Models, Development-Staging-Production Environments | MLOps Engineer | Individual Contributor (Junior to Senior) | 22-24 | null | Male |
| LightGBM | Time Series / Forecasting | Fraud, Risk | Less than a month | Amazon Web Services | Inconsistency of training and experimentation environments, Access to relevant data for training, Machine learning monitoring and observability | null | null | null | Airflow | Custom Built In-house tool | Custom Built In-house tool | Custom Built In-house tool | null | null | Financial services | 250-1,000 employees | 21-100 | 21-100 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | More than 90% | Progressive Rollouts, Development-Staging-Production Environments | Software Engineer | Individual Contributor (Junior to Senior) | 35-39 | Portugal | Male |
| Lightgbm | Recommender Systems, Search | Search, Recommender systems | Less than a month | Amazon Web Services | Inconsistency of training and experimentation environments, Machine learning monitoring and observability | Custom Built In-house tool | Custom Built In-house tool | null | Airflow | Amazon SageMaker | Custom Built In-house tool | Custom Built In-house tool | GCP / BigLake | null | Technology | 250-1,000 employees | 5-9 | 10-20 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 10% and 30% | Canary Deployments, A/B Tests for Models, Progressive Rollouts, CI/CD for continuous deployment | Machine Learning Engineer | Individual Contributor (Staff+) | 35-39 | Australia | Male |
| Sklearn | LLMs, Text / NLP (Non-LLM) | Risk, Classification | Less than a month | Amazon Web Services | Inconsistency of training and experimentation environments, Building production-grade machine learning and data pipelines, Machine learning monitoring and observability | MLFlow | null | null | Prefect | Amazon SageMaker | Sagemaker | Custom Built In-house tool | null | Amazon Bedrock | Healthcare | 50-250 employees | 100-1000 | 100-1000 | Central Machine Learning Platform / Team, AI Inventory (Keeping track of all AI usecases and models) | Between 50% and 90% | A/B Tests for Models, CI/CD for continuous deployment, Development-Staging-Production Environments | Machine Learning Engineer | Manager | 30-34 | India | Male |
| Sklearn | Search, Time Series / Forecasting | Search | Less than 6 months | Azure | Gaps in tooling and support for model productionisation, Showcasing business impact and business value, Machine learning monitoring and observability | MLFlow | null | null | Custom Built In-house tool | Databricks | Custom Built In-house tool | Custom Built In-house tool | Azure DataLake | null | Transportation & warehousing | 250-1,000 employees | 5-9 | 10-20 | Central Machine Learning Platform / Team, Data Platform / Data Engineering Organisation | Between 10% and 30% | Development-Staging-Production Environments | Machine Learning Engineer | Individual Contributor (Staff+) | 35-39 | Canada | Male |
Use the filters above each column to filter and limit table data. Advanced searches can be performed by using the following operators:
<, <=**, **>, >=, =, *, !, {, }, ||, &&, [empty], [nonempty], rgx:
TableFilter v0.7.2
©2015-2026 Max Guglielmi
Close
?
The Institute for Ethical ML