Back to Module 1.8: Vector DatabasesComing Soon
AI Lesson & Submodule
Hybrid Search
Combine sparse keyword search indexes with dense semantic vector lookups.
Why This Matters
Hybrid searches capture both exact keyword terms (product IDs, codes) and conceptual semantic queries.
What You Will Learn
- •Combine BM25 and embeddings
- •Normalize ranking scores
- •Tuning hybrid weights
Concepts Covered
Sparse BM25 indexesReciprocal Rank Fusion (RRF)Hybrid weights tuning
Mapped Foundation Project: Semantic Product Search
Search workspace loading catalog inventories into vector databases, supporting dense vector lookups and metadata query filtering.
Architecture Preview
Search gateway loading documents into vector databases, querying them on keys, and merging results streams.
Search Query InputEmbeddings TransformerVector DB Index
Tech Stack Planned
ReactTypeScriptTailwind CSS
GitHub: Coming SoonLive Demo: Coming Soon
Coming SoonTechnical Interview Value
- ?Explain how Reciprocal Rank Fusion (RRF) merges results from keyword and semantic vector search paths