What is Synapse Data Science

Definition

Synapse Data Science in Microsoft Fabric is the workload for building, training, and operationalising machine-learning models, providing notebooks, experiment tracking, and model management integrated with lakehouse data and OneLake within the unified platform.

« Back to Glossary Index

Builds and trains ML models on lakehouse data within Fabric
Tracks experiments and manages models in one integrated workload
Uses notebooks with familiar Python and ML libraries
Operationalises models alongside the rest of the data platform

Real World Example

A data scientist uses Synapse Data Science in Fabric to train a churn model on lakehouse data in a notebook, track experiment runs, and register the best model, all without leaving the unified platform.

FAQs

What is Synapse Data Science used for?

Building, training, tracking, and managing machine-learning models on Fabric data using notebooks and ML tools.

How does it access data?

It works directly with lakehouse data in OneLake, so models train on the same governed data used elsewhere in Fabric.

Does it support experiment tracking?

Yes, it includes experiment and model tracking to manage and compare training runs.