Writing
Ideas for building agentic AI systems that work in the real world.
I write about the technical and organizational systems required to operate AI at scale.
Agentic AI
Operating Models for Agentic AI Systems
Why production AI requires explicit operating models, decision rights, escalation paths, and governance.
Observability
The Five Layers of Agent Observability
Observability needs to move beyond infrastructure and model metrics when systems begin making meaningful decisions.
Decision Systems
Decision Lineage vs. Decision History
Understanding why an AI system made a decision requires more than a record of what happened.
AI Governance
Managing Risk in Agentic AI Systems
Why risk budgets provide a practical framework for determining where AI systems should operate autonomously and where humans should remain involved.