Company Agent Bench: first benchmark testing agents on internal company understanding
philhchen · x · 2026-11-13
joshzqin introduces Company Agent Bench, arguing that being good at individual tasks doesn't make an agent a good company agent — it needs deep understanding of people's priorities, judgment, and working relationships. He built the benchmark from internal Slack data, filling a gap since no company-agent benchmark existed. philhchen endorses him as a rising researcher with rare technical insight, resourcefulness, and creativity.
Related event: Company Agent Bench: Understanding People Is Enterprise Agents' Biggest Gap(4 posts)→
More from coding & agent
- Gemini API Managed Agents now support Agent Substrate to run on your own cluster — rakyll · 2026-11-13
- Using agents to reimplement MD forcefields: huge speedups possible, but forces must still agree — rishabh16_ · 2026-11-13
- Fish Audio launches Fish Agent, a no-code platform for AI voice agents that make calls — Shruti_0810 · 2026-11-13
- With AI help, a team finally migrated all their infra to K8s after ragequitting years ago — DanielLockyer · 2026-11-13
- Agents trust other agents by default — one user's skeptical-agent fix for a real security hole — andrey_kurenkov · 2026-11-13
- A dead-simple canary test for drifting agent sessions: make it address you by name — brandon_galang · 2026-11-13