Writing

Notes on AI systems, machine learning, and the work around them.

Project breakdowns, technical lessons, wrong turns, and occasional notes on work and life.

002Systems notes8 min read

Building a self-hosted multi-agent runtime

Why I built my own agent system around the Claude Agent SDK, and what memory, specialist agents, recovery, and trust changed along the way.

Read note
001Model notes5 min read

What ROC-AUC hid in my fraud-detection project

A high score told me the model could rank transactions. It did not tell me how many alerts a fraud team would have to review.

Read note

What belongs here

I write when a project changes how I think, build, or work.