Now · updated 4 Oct 2026
What I’m doing now.
Building evaluation and reliability tooling for agentic AI, and writing about what assurance should mean for autonomous systems.
A now page is a snapshot of current focus. It changes as the work does.
Focus
- Architecting evaluation, reliability and observability for agentic AI systems in production.
- Strengthening AI security practice for agents: prompt injection, tool and MCP security, excessive agency.
- Shaping AI governance for enterprise deployments, aligned to NIST AI RMF and regional regulation.
Publishing
- Turning field notes on evaluation and governance into published essays on this site.
Open to
- Advisory and architecture reviews for agentic AI assurance and security.
- Speaking on evaluation, red-teaming and AI governance.
- Senior leadership conversations.