Writing

Thoughts on AI, engineering, and the things I'm building.

May 1, 2026·10 min read

The Column Named amt_v2 Is Why Your Enterprise AI Fails

Enterprise AI systems don't fail because the models are bad. They fail because the context is wrong. Here's how I built a multi-agent pipeline to fix that — and the design decisions that made it actually work.

Read article
April 28, 2026·8 min read

The AI Agent Harness: Why Your LLM Needs a Control Layer

Large language models are powerful but chaotic on their own. An agent harness is the scaffolding that turns raw LLM intelligence into reliable, goal-directed software. Here's how it works — and why it matters.

Read article