---
title: "Hybrid retrieval launch checklist: a Monday checklist"
url: https://memory.wiki/f7pTxchi
updated: 2026-11-22T21:52:01.793Z
source: "memory.wiki"
---
# Hybrid retrieval launch checklist: a Monday checklist

## Why "Hybrid retrieval launch checklist" matters for local RAG

BM25 + vectors + eval queries.  Retrieval-augmented generation is usually the first serious project a team builds on a local model, because it keeps documents in-house and avoids fine-tuning. It is also where small design choices decide whether people trust the answers.

## Ingestion comes first

Bad input produces confident wrong answers. Before thinking about embeddings:

- Extract structure: headings, tables and lists should survive conversion from PDF or Office files.
- Attach metadata to every chunk: source file, section, version and document date.
- Reject documents that have no owner or no date; stale policies are a liability.
- Keep the original files so you can re-ingest when your splitter improves.

## Chunking and retrieval

Start with moderate chunk sizes and some overlap, then evaluate on your own questions instead of copying settings from a blog. Chinese text needs a splitter that respects sentences rather than English whitespace rules. Pure vector search misses exact identifiers such as part numbers, contract IDs or SKUs, so keep a keyword index alongside the vector store and merge the results. When policies change over time, bias retrieval toward the newest valid version.

## Evaluation you can repeat

Build a set of real questions with known answers and the documents that should support them. After each change to chunking, embedding model or prompt, run the set and look at both the answer and the retrieved passages. If the right passage was not retrieved, no prompt will fix it.

## Operations

- Back up the vector store and test a restore.
- Separate collections per department or client so access rules are easy to enforce.
- Log which sources were cited for each answer.
- Decide what the assistant must refuse to answer, and publish that list.

## Questions for your pilot users

- Did the answer cite the document you expected, and was it the current version?
- Was anything important missing that you know exists in the documents?
- Did the assistant refuse when it should have, and answer when it should have?
- How long did it take to verify the answer using the cited source?

Collect these answers in a simple form during the pilot. Patterns in the feedback usually point to ingestion or retrieval problems, which are cheaper to fix than switching models.

More on this topic: [local RAG knowledge-base guide](https://microaione.com/tutorials/local-rag-knowledge-base/). See also [MicroAI One software](https://microaione.com/microai-one/).

## Bottom line

Local AI rewards patience and record keeping more than raw spending. Decide what the system must do, measure it on your own hardware and documents, write down every change with a date and an owner, and review the setup after each significant upgrade. Small, documented steps are easier to debug, easier to hand over and easier to defend when someone asks why a particular choice was made. When in doubt, test on a smaller scale first and expand only when the evidence supports it.

## Quick reminders

- Pin driver and runtime versions together.
- Leave memory headroom for the OS and the embedding model.
- Use one models directory shared by every runtime.
- Publish what the assistant will refuse to do.

*Published by [MicroAI One (本地AI 1号网)](https://microaione.com/), a Shenzhen team working on local LLM hardware, software and deployment.*

---

## Summary
Successful local RAG deployment relies on high quality data ingestion, hybrid keyword and vector retrieval, and rigorous evaluation using a set of known questions. Small, documented changes and systematic testing are more effective for building reliable systems than simply increasing model scale.

## Themes
- Local RAG deployment
- Hybrid retrieval strategies
- Systematic evaluation methods
- Data ingestion best practices

## Key takeaways
- Hybrid retrieval combining BM25 and vector search is necessary for accurate document identification.
- Metadata including source, section, and date must be attached to every chunk to ensure data integrity.
- Evaluation must be performed using a set of real questions with known answers rather than relying on generic settings.
- Access control should be managed by separating vector store collections by department or client.
- System performance should be tested on a smaller scale before expanding to ensure evidence supports the configuration.

## Insights
- The document suggests that retrieval failures are often misattributed to prompt engineering when they are actually caused by poor chunking or ingestion.
- Operational stability is prioritized over model performance, emphasizing that record keeping and versioning are more critical than raw hardware spending.
- The document advocates for a hybrid search approach specifically to solve the failure of vector models to retrieve exact identifiers like SKUs or part numbers.

## Open questions / gaps
- The document does not specify which exact embedding models or vector databases are recommended for these hybrid setups.

