API Governance in the age of AI
summary
Microsoft Graph platform documents 12,000+ resource types, 2,000+ relationships, 100 product teams, three review boards and about 2,500 API reviews a year; reviewers spend ~25% of time on one‑on‑one education. Governance uses Azure DevOps plus automated rules for permissions, docs and security, and experiments apply retrieval augmented generation with Azure AI Search, Azure OpenAI, blob storage, the a002 embeddings and semantic kernel to ground responses. Results show near‑100% relevance for simple queries, a measured ~20% faster reviews and up to ~40% automation potential, with humans kept in the loop for complex cases.
AI-written summary · about
chapters (6)
- 0:00API governance introduction
- 3:39Microsoft Graph API scale
- 7:13API review process
- 11:58AI semantic search for governance
- 16:18AI governance results and future
- 17:43Q&A
related talks
26mMastering API Governance: Microsoft's Blueprint for Success
13mAI from Microsoft's Perspective
22mMaking AI Accountable: Azure AI Foundry
23mAI assisted code reviews Quality and Security at Scale
15mAI Guardrails for Enterprise Deployments
31mAzure AI Innovations at Microsoft
30mClean and Controlled AI with API Management
50mAzure AI, Power new possibilities for every organization