Integrating Azure AI and Fabric for Next-Gen AI Solutions
summary
Claims that integrating Microsoft Fabric and Azure AI lets teams use OneLake data to build RAG-based assistants with lower friction, backed by an IDC ROI stat of $1→$3.5 in ~14 months. Practical pieces shown: creating an AI Skill in Fabric that publishes an endpoint, pushing Unstructured and structured files into a vector index in Azure AI Search, and composing retrieval+generation in Azure AI Studio with Prompt flow. Demoed Contoso home-banking assistant, embedding Azure OpenAI models and evaluation that raises relevance nearly a point, proving the pipeline works end to end.
AI-written summary · about
chapters (11)
- 0:00Welcome
- 1:16GenAI investments ROI
- 6:20Data strategy for AI
- 8:52Customer use cases
- 10:15Fabric AI scenarios
- 13:24Copilot in Fabric
- 17:38AI Skills in Fabric
- 22:02RAG assistants with Azure AI
- 33:32Prompt flow for Copilot
- 41:01Evaluating AI assistants
- 43:38Summary
related talks
26mTransform productivity with AI experiences in Microsoft Fabric
20mUnlocking Azure OpenAI: A 20-Minute Introduction
17mAzure Cosmos DB in AI Workflows: From Knowledge Retrieval to Microsoft Fabric Integrations
31mAzure AI Innovations at Microsoft
45mHarnessing Data Science and AI in Fabric
47mData analytics for the era of AI
01h 05mKeynote: Building Applications with the Modern SQL Using AI and Microsoft Fabric
23mBuilding AI-Driven Real-Time Lakehouse APIs