Writing archive

Writing on AI systems in practice

Eleven notes on agent architecture, governance, reliability, model economics, and professional development.

Sep 12, 20262 min read

A reflection on building and deploying AI in a high-stakes product environment before beginning a new venture.

Sep 3, 20262 min read

Anthropic’s advanced certification for architects designing and governing Claude solutions at enterprise scale.

Aug 24, 20263 min read

As agent systems become active web users, products need explicit identity, permissions, interfaces, and evidence trails.

Aug 3, 20261 min read

A six-week partner cohort, hands-on labs, technical reviews, and practical infrastructure work strengthened my multi-cloud foundation.

Jul 29, 20264 min read

AI lowers the barrier to persistent, adaptive automation—and raises the reliability demands placed on the systems underneath it.

Jul 28, 20262 min read

I earned GitHub's Agentic AI Developer credential after taking the exam during its beta period.

Jul 23, 20263 min read

A multi-agent system creates risks through the sequence of actions its participants can take together.

Jul 11, 20263 min read

One real coding task cost roughly 80 times more with one frontier model than another, with negligible difference in the final result.

Jun 27, 20263 min read

A system should grant agents more room to act only when risk, evidence, and reversibility support it.

Jun 27, 20264 min read

Event-driven architecture gives AI agents a cleaner way to coordinate, fail, recover, and evolve inside real enterprise platforms.

Jun 27, 20263 min read

Cursor, Claude, Copilot, and custom agents are most valuable when they become part of a governed engineering workflow instead of a side tool.