How better search took an AI agent from 36% to 98% accuracy
AI Engineer · youtube · 2026-11-03
Aamir Shakir of Mixedbread argues agents are usually smart enough — the real bottleneck is data access. The typical 'Frankenstein stack' of OCR, chunking, embeddings and a vector DB loses information and forces agents to read every document. He demos multimodal, late-interaction retrieval on Intel earnings reports and shares a customer whose answer accuracy jumped from 36% to 98% after switching search.
More from coding & agent
- Running a family trip with two agents: Teamily AI in action — thetripathi58 · 2026-11-03
- Dev says AI coding is collapsing to intent plus privileged SOR access — curious_vii · 2026-11-03
- 开发者开源 Cosmion：让 Claude Code 等 Agent 一键生成图片 — TheAverageJedi · 2026-11-03
- ComfyUI 新节点按语义分段 conditioning，缓解 H3 长视频循环 — Commercial-Ad-3345 · 2026-11-03
- Designer builds fully functional websites from Figma frames in a day using Claude — Tegadesigns · 2026-11-03
- AI Slop Addons Flood World of Warcraft, Forcing Warcraft Logs to Rework Its API — 404 Media · 2026-11-03