Turbo charge your Data Science workflow with Microsoft Fabric
summary
Claim: Microsoft Fabric supports an end‑to‑end data science workflow inside one SaaS. Builds two data products on the Titanic dataset: a diagnostic semantic model (medallion bronze→silver→gold) and a predictive pipeline that trains with scikit‑learn, logs experiments to MLflow, validates inputs with Great Expectations, and connects notebooks to Power BI via Semantic Link (Senpai). Demo runs multiple model versions, selects version 12 (~80% accuracy) for batch deployment, writes predictions back to OneLake, and alerts data stewards via Teams on validation failures.
AI-written summary · about
chapters (18)
- 0:00Introduction
- 1:52Data as socio-technical
- 3:15Product mindset in data teams
- 5:00Titanic data science demo
- 7:47Medallion architecture
- 9:15Validating data with Great Expectations
- 13:51Semantic models for analytics
- 15:37Power BI reports
- 18:24Predictive analytics with semantic link
- 20:35Exploratory data analysis with Senpai
- 26:43MLflow model training
- 31:41Collaborative model tuning
- 33:48Deploying batch models
- 36:50Batch prediction results
- 38:55Fabric vs Azure ML
- 43:55Data product lifecycle
- 45:20Accelerating value flow
- 46:22Q&A
related talks
22mModels to outcomes with end-to-end data science workflows in Fabric
26mTransform productivity with AI experiences in Microsoft Fabric
47mData analytics for the era of AI
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
21mDataEngBytes 2023 - SYD-T1-06 - Sana Sanai
37mPracticing Data Science in the cloud