Turing launches CEO Bench: 500+ expert tasks test frontier agents in a simulated company
ecekamar · x · 2026-09-10
Turing introduced CEO Bench, an eval suite testing whether frontier agents can do long-horizon financial and operational work inside a complete fictional company.
- First virtual company, Cirrus Sleep, Inc., is a simulated DTC sleep brand with 1100+ files: accounting, finance, contracts, HR/payroll, board materials, ops, marketing, support, and a year of internal Slack
- 500+ human-authored executive/department-level tasks, 40 published and scored in this release, all written by finance and operations practitioners
- Moves from gold answers to graded judgment, with human experts in the loop and agentic QC per task
- Motivated by benchmark contamination and the gap between saturated public benchmarks and real deployment failures
Built on their SVC (Simulated Virtual Company) platform; available now.
Related event: Turing Launches CEO Bench for Enterprise AI Agents(2 posts)→
More from coding & agent
- SF 'Code Engineering' event picks up where 1,000-person harness engineering meetup left off — LangChain · 2026-10-27
- Prompt turns Claude Code into an iPhone backup forensics tool exposing 30-day 'deleted' photo copies — alex_verem · 2026-10-27
- Prompt turns Claude Code into a forensics tool finding deleted iPhone photos in backups — alex_verem · 2026-10-27
- Phones won't be mobile IDEs but control planes for agent fleets, argues Yuchen Jin — Yuchenj_UW · 2026-10-27
- FrontierCS 2 launches on Kaggle: 48 tasks make agents beat published CS papers — MeganRisdal · 2026-10-27
- Agent Field Kit launches $29 security runbooks for production AI agents — alvelda · 2026-10-27