Model Architectures · Practitioner
Grounding with Azure AI Search: Vector, Hybrid, Semantic and Agentic Retrieval
Azure AI Search is the retrieval engine behind almost every serious Azure RAG system. Learn the four retrieval modes, how to chunk, and how to keep one user out of another's documents.
- 1.1 hours lesson
- intermediate level
- Azure AI Track
- Narrated presentation
What you’ll learn
- Keyword BM25 for exact identifiers, vector for paraphrase, hybrid with RRF as default
- Semantic ranker as a re-ranking layer with captions and extractive answers
- Agentic retrieval: query decomposition over a conversation, parallel subqueries, unified ranked result
- Chunk on document structure and carry source, page and heading fields for real citations
- Integrated vectorization removes the separate embedding pipeline
- Security trimming with Entra ID group filters - never in the prompt
- AI Search bills hourly for provisioned capacity, plus per-query semantic ranking
Inside this lesson
A narrated, slide-by-slide presentation with 1 prediction exercise, 1 knowledge check, 3 videos. You answer as you go, and the platform checks your work.