Writing · 14 pieces
Notes from
the work itself.
Essays and field notes on software architecture, AI-enabled engineering, developer experience and business trade-offs. Published here in full and kept independent from social platforms.
Browse the archiveSelected writing
Release noteLinkedIn post
Peace of mind comes from verification
A private Cloudflare uptime monitor, Telegram alerts, and the live checks that turn an AI-built project into something dependable.
Release noteLinkedIn post
Building the baby sleep app our family wanted
What daily use of Napper taught me about the freedom and maintenance of owning a custom family app.
Field noteLinkedIn post
Architecture first. Implementation matters.
The connections between architecture, AI, business strategy and engineering judgement that this website exists to explore.
Build noteLinkedIn post
Building and deploying an app from my phone
A small product experiment in which the constraints are part of the implementation story.
AI in practiceLinkedIn post
Kimi K3 and the quality of collaboration
A week of using Kimi K3 prompted a different question about AI assistants: how pleasant are they to work with?
AI in practiceLinkedIn post
A few AI-related thoughts I had lately
On stopping model drift early, the blast radius of vibe coding, supervising multi-agent work and building software that does not use AI at runtime.
AI in practiceLinkedIn post
The reward loop of AI coding tools
Big wins, small rewards and near misses: a reflection on the pull of just one more prompt.
AI in practiceLinkedIn post
Harness engineering
The tools, context, verification and approval gates around a model determine what an agent can reliably do.
Event noteLinkedIn post
Three takeaways from TypeScript AI Demo Day
Harness engineering, agent skills and security, and the changing role of user interfaces.
AI in practiceLinkedIn post
AI supervision in infrastructure and security
Why the consequences of a plausible but wrong answer make infrastructure and security a different supervision problem.
AI in practiceLinkedIn post
How long will AI supervision stay critical?
A dated prediction about the pace of AI progress and why learning to supervise well matters now.
Engineering cultureLinkedIn article
We Know. We’re Just Not Saying It.
A note on the things engineering teams often see clearly before they are ready to say them out loud.
Engineering practiceLinkedIn article
4 takeaways from The Unicorn Project
Notes on developer tools, technical debt, trust, teamwork and making room for experimentation.
AI in practiceLinkedIn article
What we learned from our journey to AI
Lessons from building recruitment automation with transparency, responsible use and human judgment in mind.