Live session
RS
Enterprise Knowledge Assistant

Week 2 · Academic Concept Studio

Chunking, embeddings, retrieval and grounded generation

29 Sept 2026, 17:30 GMT+5:3090 minutes · IST (Asia/Kolkata)Dr Ananya Rao
Ended

Session recap

Covered: tokens, context limits and why retrieval is needed; chunking and metadata experiment; embeddings and semantic similarity.

  1. Recording

    Ready to watch.

  2. Check my understanding

    Explain and apply the concepts

    Start check
  3. Next project task

    Concept exercises and a working retrieval component.

    Open workspace

What you’ll build or understand

  • Explain how chunking affects retrieval
  • Describe what an embedding represents
  • Build a grounded prompt from retrieved passages

Agenda

  1. 20 minTokens, context limits and why retrieval is needed
  2. 25 minChunking and metadata experiment
  3. 25 minEmbeddings and semantic similarity
  4. 20 minConstructing grounded model inputs
Demo workspace · Sample data