AI-native CPTO. I run product and engineering at Darwin with a small team and a fleet of coding agents, and I still ship code every day.
| Last 30 days | |
|---|---|
| Pull requests from my account, most written with agents | 638 |
| Merged | 568 (89%) |
| Contributions, last 12 months | 9,759 |
| People on product and tech | 5 |
Updated 2026-10-12.
Most of this work lives in private repositories. The method behind it is public.
- ai-native-cpto: the operating model. Autonomy by default, the checkpoint after the work, owners who rule on the diff, and a measure behind every rule.
- agent-guardrails: hooks that gate agent actions by tier. An irreversible action needs a human in front of the screen.
- cpto-skills: the skills and commands my agents run, from spec to pull request.
Darwin is a nature risk intelligence platform. It measures how companies depend on and impact nature and climate across their sites and supply chains, and turns that into TNFD and CSRD reporting. Stack: Rust/Axum, Vue 3/TypeScript, PostgreSQL/PostGIS, AWS, DDD and hexagonal architecture.
- Co-founder and CPTO, Darwin (2024 to now)
- Investment Manager, Samaipata VC (3.5 years)
- Strategy consultant, L.E.K. Consulting (2 years)
- Embedded software engineer, Parrot, drones (1 year)
- MSc in Engineering (CS), Management and Finance: HEI, Bocconi, ESSEC
- Essays on running engineering with agents, one new essay most weeks
- The Finch, newsletter on nature and data
LinkedIn · X · Bluesky · darwindata.ai



