Writing · Medium since 2018
Writing
Notes from building AI agents, moving teams onto AI workflows and automating operations in regulated industries. Earlier posts cover scalable infrastructure, GraphQL and React. Twenty-plus posts since 2018.
Currently trackingEncoded-judgment skills · the audit-trail problem in regulated AI · how PE operating teams should think about model-vs-prompt leverage · what makes a Claude Code skill citable.
Newsletter
Get new pieces on building AI agents, moving teams onto AI workflows and automating operations by email when they come out.
Free. Unsubscribe any time.
2026 — agentic AI
2026
-
Sep 24
The on-call engineer who never sleeps: building an AI agent to triage FIX trading incidents · Medium
-
Aug 18
AI due diligence: how to assess an AI company before you invest
-
Aug 18
Is the AI real, or a wrapper? A technical due diligence guide
-
Aug 18
How to audit AI agents: a practical playbook for regulated firms
-
Jul 28
Your AI agent works in the demo. Can you explain what it did last night? · Medium
-
Jun 10
I built a self-debugging AI agent from scratch - here's exactly how the reasoning cycle works · Medium
-
Jun 04
I Review Pull Requests With AI From My Terminal — Here's How to Set It Up in 15 Minutes · Medium
- May 11
-
May 04
I built a Claude Code skill that does a 20-minute Azure cost review · Medium
-
Apr 29
From markdown notes to a sized Jira backlog in five minutes · Medium
-
~May 06
The model-vs-prompt split, and why portfolio operations sits on the prompt side · coming
Earlier — architecture & engineering
2022
2021
- Nov 14
- Oct 24
- Oct 16
- Oct 09
- Apr 17
- Apr 10
- Apr 03
- Mar 29
- Mar 20
- Mar 05
- Feb 26
- Feb 26
- Feb 25
- Jan 13
- Jan 13
2018