I only write about what I've shipped.
Platforms, teams, hiring, and where AI actually adds value. Some pieces start here; others are mirrored from where they first ran.
The AI Stack Trifecta
Automation, data, and application layers, and why AI agents need all three to work.
Engineers Are Optimizing for the Wrong Kind of Correct
AI isn’t producing slop. Your missing quality gates are.
We Looked at Claude Code's Architecture. It Looked Familiar.
What Anthropic's leaked source confirms about building agentic systems for the enterprise.
Democratizing Enterprise Machine Learning with AutoML
How we reduced ML model build time by 95% and enabled domain experts to deploy production-grade models in days instead of weeks.
4 pieces