15 years in industrial quality. Building practical software with coding agents.
My background is in electronics, automation, supplier quality, and structured problem solving.
I usually define the problem, constraints, and acceptance criteria, then use coding agents to help implement and verify the work.
A lot of what I build starts from the same question: can this remove real friction or make a difficult process easier to understand, use, or validate?
Right now I'm working across a few different areas.
Quality-Engineering-Skills is a collection of reusable AI skills for 8D, VDA, NCR, audits, root-cause analysis, and supplier quality.
In my day-to-day quality work I also explore and apply practical ways AI can support Quality Management.
A Team and the builder.* projects are experiments around coding agents, reusable skills, and more structured ways of turning an idea into something that can actually be tested.
I've also been spending a lot of time exploring smart glasses and wearable computing.
Agent HUD explores how an AI agent can ask for a human decision without constantly pulling someone back to a laptop or phone. It started with Raven Prism and has also been ported to Brilliant Labs Halo.
DOOMed Prism started as a slightly unreasonable idea — running DOOM on lightweight smart glasses — and became a useful way to explore rendering, input, voice interaction, simulation, and real hardware constraints.
I'm also researching persistent, local-first memory for assistants, agents, physical projects, and wearable computers.
I tend to move between software, hardware, AI, and quality engineering rather than staying inside one layer.
The process is usually similar:
understand the problem → define the constraints → build → test → learn from what actually happens