How a $6B Hedge Fund Measures AI ROI With Live A/B Tests and Cost per Useful Insight
AI Engineer · youtube · 2026-10-24
At the AI Engineer summit, Hari Kumar, head of AI at Contour Asset Management (a $6B fundamental long/short fund), explained how the fund quantifies returns on AI spending:
- Live A/B testing of analysts' research with and without AI tools, measuring the process rather than just P&L
- A homegrown metric: "cost per useful insight," plus time-saved tracking to break AI gains into computable units
- Idea-resurfacing agents that monitor old research ideas and flag them when market conditions make them relevant again
- Rollout across trading, ops, and compliance, and squeezing more value from existing alt-data spend
He also flagged a commonly overlooked hidden cost: maintaining and fixing the AI tools themselves.
More from Companies & People
- Winning NVIDIA's Build-a-Claw Paris hackathon — and keynoting two meters from Jensen Huang — lifebypixels · 2026-10-24
- Commentary reads Kimi's KOL blitz as cash crunch: only two survival paths left in AI? — teortaxesTex · 2026-10-24
- Pedro Domingos: Sam Altman's AI race strategy is cornering the compute market — pmddomingos · 2026-10-24
- Raspberry Pi CEO Eben Upton on building a textbook-priced computer that sold 75M units — sarahdrinkwater · 2026-10-24
- The phone is the computer: how device makers and model labs battle over personal agents — sujingshen · 2026-10-24
- The 'Hugging Face mafia' map: ex-employees' startup empire, explorable for any company — TheZachMueller · 2026-10-24