Models to outcomes with end-to-end data science workflows in Fabric
summary
Microsoft Fabric combines Power BI Data Factory and Synapse into a single platform and shows an end to end data science workflow that actually works from exploration to deployment. It demos cleaning and feature engineering with Data Wrangler, training distributed LightGBM on Spark over 34 million records, tracking runs and versions with MLflow auto logging, batch scoring via SynapseML and seamless consumption in Power BI using Direct Lake mode, plus Semantic Link/SemPy access to measures and planned Azure OpenAI copilots.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 0:25Agenda
- 1:12Why data science is so valuable
- 4:00Data science workflow
- 4:29Data science process
- 6:19Data Wrangler
- 6:53MLflow
- 7:41Guided experience
- 8:00synapse ML
- 8:30core capabilities
- 9:03language support
- 9:26Spark R
- 9:46Notebook experience
- 14:24Roadmap
- 15:40Semantic Link
- 16:26Demo
- 19:15AI Cognitive Services
- 19:37Developers CoPilot Experiences
- 20:22DataCentric AI Plugins
related talks
51mTurbo charge your Data Science workflow with Microsoft Fabric
45mHarnessing Data Science and AI in Fabric
20mGetting started with SparkML in Microsoft Fabric
52mPowering Up Your Power BI Development Workflow with Microsoft Fabric Notebooks
49mYour first flight with Data Factory in Microsoft Fabric
50mFabric: What's new and what's next
46mIntegrating Azure AI and Fabric for Next-Gen AI Solutions
14mUsing Microsoft Fabric Workload for AuraDB to Make Product Recommendations from OneLake Data