From Retrieval to RAG
Chunking, filtering, hybrid search, RRF, reranking, and evaluation
THE QUESTION THIS PAGE ANSWERS
ANSWER FIRSTWhat is the key idea behind “From Retrieval to RAG”?
Chunking, filtering, hybrid search, RRF, reranking, and evaluation
Make the claim earn its place. Use this page as a decision aid, not a definition to memorize. Connect the idea to one real task, one observable result, and one failure that would change your mind.
Write one question you could answer with evidence after trying this idea.
A conclusion that sounds complete but leaves the key assumption untested.
Embed query
Use the same model as ingestion.
Retrieve
Tenant/ACL filter plus Top-K search.
Rerank
Rerank, deduplicate, and budget tokens.
Generate
Answer only from evidence and cite sources.
Chunking
Split on titles, paragraphs, and semantic boundaries with modest overlap. Large chunks add noise; tiny chunks lose context.
Metadata
Keep source ID, version, section, tenant, ACL, and update time for citations, filters, and deletion.
Versioning
Re-embed when content or the model changes. Blue/green collections avoid mixing incompatible vectors.
| Method | Purpose | Caution |
|---|---|---|
| Scalar filter | Tenant, ACL, time, language | Enforce access during retrieval, never only in the prompt |
| Vector + keyword hybrid | Semantics plus exact SKUs/error codes | Raw score scales are not directly comparable |
| RRF | Fuse several rankings | Simple and stable; rerank the resulting candidates |
| Partition / TTL | Hot-data isolation or expiration | Advanced options; start with clean collections and filters |
Why “Embed query” can find relevant content
“Tenant/ACL filter plus Top-K search” moves retrieval beyond storing material: the real question is how to find what is relevant. That decision shapes the input quality of RAG, recommendation, and image-search systems.
Similarity is not the answer
In the flow described by “Rerank, deduplicate, and budget tokens”, embeddings place items in a comparable semantic space and a neighbor index narrows the search. The final answer still depends on whether the retrieved chunks cover the question, whether the distance metric fits, and whether the evidence is current.
Separate findable from relevant
Turn “Re-embed when content or the model changes.” into a small test: prepare queries with known answers, record relevance, misses, and distractors, then decide whether chunking, the index, or reranking needs to change.
From “Embed query” to “Retrieve”
“Embed query” grounds the problem in “Use the same model as ingestion”. “Retrieve” then moves it toward “Tenant/ACL filter plus Top-K search”. Together, they show that the lesson is not just a conclusion to remember, but a claim with conditions.
Carry the judgment into the next situation
The same logic applies to retrieval: define what counts as relevant, check whether recall covers the question, and then inspect whether ranking, chunking, or freshness pushed useful evidence out.
- “Embed query”: Use the same model as ingestion
- “Retrieve”: Tenant/ACL filter plus Top-K search
- “The closing point”: Re-embed when content or the model changes. Blue/green collections avoid mixing incompatible vectors
The final “The closing point” brings the discussion to “Re-embed when content or the model changes. Blue/green collections avoid mixing incompatible vectors”. The useful thing to carry forward is knowing which judgments must be revisited when input, scale, or risk changes.
I turned one judgment from this article into a small experiment I could run today. Knowing what to observe next is more useful than simply remembering the conclusion.
After reading this, I first looked for the conditions behind the idea instead of copying the method into a project. That order made the later trade-offs much clearer.
When this judgment reaches real work, which constraint should be added first? I am curious which step matters most between reading and the first practical attempt.
No discussion on this article yet.