Challenges in building high-recall RAG: balancing coverage, reliability, and cost
iMiguelmars · reddit · 2026-09-01
The author discusses building a system that turns large, messy archives into verifiable memory, aiming to detect duplicates, contradictions, superseded info, and provenance while maintaining high recall. The core tradeoff involves balancing coverage, reliability, and cost. Current experiments include sliced reading, independent review stages, mechanical validation, caching, and long-running workflows. The author seeks insights on reducing costs without sacrificing recall, representing contradictions/provenance, and architectural scalability.
Related event: Building a High-Recall, Traceable "Second Brain" RAG System(2 posts)→
More from coding & agent
- Using Grok Bot to build college admissions dataset pipeline — lennysan · 2026-09-01
- Idea: 'Money Leak Hunter' Grok Bot for finance audit — lennysan · 2026-09-01
- From Discord Bots to a Multiplayer Agent Workspace — steipete · 2026-09-01
- OpenClaw puts the agent on your machine: 933 volunteers vs rented AI assistants — heyneighbor · 2026-09-01
- Dev shares a dirt-cheap approach to visual diffs — zeeg · 2026-09-01
- Grok Bot automates Shopify updates and supplier coordination — billyjhowell · 2026-09-01