I design AI-assisted systems that separate capability from authority, preserve human decision rights, and return inspectable evidence instead of asking for blind trust.
Contact: secondmindsystems@gmail.com · Evidence portfolio
- Governed Change Demo — a public, deterministic reference implementation of governed change. Run it in five minutes.
- Governed AI Systems Portfolio — case studies, evidence, limitations, and claim boundaries.
- Second Mind Systems — the company front door and product map.
- A proposed change can be checked against explicit policy before it is accepted.
- A blocked change can be repaired and evaluated again.
- Decisions and replay results are preserved as inspectable receipts.
- 76 automated tests pass in the current published validation suite.
The public repositories demonstrate bounded prototypes, documented methods, and reproduced local results. They do not establish cryptographic security, production deployment, universal agent conformance, compliance certification, or access to private Second Mind runtime machinery. Independent separate-machine reproduction remains an open evidence objective.
I define the architecture, authority boundaries, and acceptance decisions. AI agents may assist with bounded implementation and review; their output is checked against repository truth and explicit claim limits.
Explore Second Mind Systems · View the portfolio · Run the demo