Fifteen-ish years in. Started with Java and Eclipse and swiftly moved to .NET, drifted into mobile (Swift, Flutter), spent the last few years writing a lot of Go and TypeScript and trying to keep teams pointed in roughly the same direction.
Currently leading engineering at AJ Hackett Bungy. These days a lot of my work is AI-shaped: building agents, wiring up MCP servers, figuring out where this stuff actually earns its keep versus where it's just noise. The rest is the usual — architecture, code review, repeat.
A small corner of the internet for things I've built, things I'm thinking about, and things I'll probably change my mind on later. It’s plain on purpose.
A tiny macOS menu bar app that watches GitHub Actions across every repo you can see. One glyph for the worst current state, ETag-aware polling, notifications the moment something turns red. Built because I was tired of keeping a tab open on Actions all day.
A local-first multi-agent orchestration platform. React + Vite up front, Express and SQLite at the back, one control surface for missions, agents, issues, and run history. The kind of thing I kept rebuilding badly until I sat down and built it properly.
A self-hosted AI operations agent in Go. Embedded SQLite, pluggable LLM router, connectors for Microsoft 365, Azure, and Slack, with an audit trail because nobody trusts an agent that can't show its working.
DMs open on x / @nzmrldev, or reach out on linkedin / in/mrlist.