Fix Bad RAG Queries Without Re-Indexing
Users ask vague questions, but technical documents use specific vocabulary. That semantic gap is why retrieval misses the right chunks. This video shows three ways to fix that without re-indexing: query rewriting (+43% score improvement), decomposition (surfaces docs direct search misses), and HyDE.
📚 This is Module 3 of a 10-part RAG course. New modules dropping regularly.
⏳ Chapters:
00:00 The Semantic Gap
00:52 Three Techniques
01:51 Code Setup
02:29 Query Rewriting
04:18 Query Decomposition
05:54 Automatic Routing
06:45 HyDE (Bonus)
08:29 Recap & What's Next
🔗 Links:
Playlist: https://www.youtube.com/playlist?list=PL0G6--HT7Yq_sxLFyWFWL6KHYWlosspj_
Discord (Get the Notebook): https://discord.gg/KpnJQbgpjt
Jupiter (AI Job Search): https://jupiterpath.dev
▶️ Next — Module 4 (Reranking): https://www.youtube.com/watch?v=qoHjx1vizak
Watch on YouTube ↗
(saves to browser)
Sign in to unlock AI tutor explanation · ⚡30
Related AI Lessons
⚡
⚡
⚡
⚡
Why StarRocks Is Better Than Elasticsearch for RAG and AI-Powered Vector Search Analytics
Medium · LLM
Production RAG: Shipping a RAG System Into an Enterprise Product
Medium · RAG
HyDE: Search With the Answer You Wish You Had
Medium · RAG
Hierarchical Indices: Find the Section First, Then Find the Sentence
Medium · RAG
Chapters (8)
The Semantic Gap
0:52
Three Techniques
1:51
Code Setup
2:29
Query Rewriting
4:18
Query Decomposition
5:54
Automatic Routing
6:45
HyDE (Bonus)
8:29
Recap & What's Next
🎓
Tutor Explanation
DeepCamp AI