🤖 Ghostwritten by Claude · Curated by Tom Hundley

This article was written by Claude and curated for publication by Tom Hundley.

---

Hybrid Search: Combining Vector and Keyword Retrieval

Why Smart search isnt always better than Dumb search.

The Exact Match Failure

Vector search is fuzzy.

• Query: Error Code 503
• Vector Result: Server Unavailable (Correct meaning, but maybe you wanted the documentation for code 503 specifically).
• Vector Result: Error Code 404 (Mathematically close, but wrong).

Keyword search (BM25) is exact.

• Query: Error Code 503
• Result: Doc: Error Code 503 Handling (Exact match).

Production systems need BOTH.

The Algorithm: Reciprocal Rank Fusion (RRF)

How do you combine a list of 10 Keyword results and 10 Vector results? They have different scores (BM25 score vs Cosine Similarity).

RRF ignores the scores and looks at the Rank.

• If Doc A is #1 in Keyword and #1 in Vector - Final Score: High.
• If Doc B is #10 in Keyword and #50 in Vector - Final Score: Low.

Sparse Vectors (SPLADE)

A newer approach is Sparse Vectors. Instead of just storing keywords, we store Learned Weights for keywords.

• It learns that in the query Apple Banana, Apple is important but Banana is not (hypothetically).
• Tools like Splade give you the precision of keywords with the learning of AI.

Implementation Pattern

1. Ingest: Save text to BM25 index and Vector index.
2. Query: Run both searches in parallel.
3. Fuse: Merge using RRF (alpha=0.5 usually works best).
4. Return: The sorted list.

Conclusion

Dont choose between Keyword and Vector. The answer is Yes.

---

*This concludes the Vector Search Retrieval series.*

How This Article Was Made

This article is a live example of the AI-enabled content workflow we build for clients.

Stage	Who	What
Research	Claude Opus 4.5	Analyzed current industry data, studies, and expert sources
Curation	Tom Hundley	Directed focus, validated relevance, ensured strategic alignment
Drafting	Claude Opus 4.5	Synthesized research into structured narrative
Fact-Check	Human + AI	All statistics linked to original sources below
Editorial	Tom Hundley	Final review for accuracy, tone, and value

The result: Research-backed content in a fraction of the time, with full transparency and human accountability.

Why We Work This Way

Were an AI enablement company. It would be strange if we didnt use AI to create content. But more importantly, we believe the future of professional content isnt AI vs. Human—its AI amplifying human expertise.

Every article we publish demonstrates the same workflow we help clients implement: AI handles the heavy lifting of research and drafting, humans provide direction, judgment, and accountability.

Want to build this capability for your team? Lets talk about AI enablement →

---