About
“We have no future because our present is too volatile. We have only risk management. The spinning of the given moment’s scenarios. Pattern recognition” - William Gibson
I see patterns in things that don’t obviously connect. It’s gotten me into trouble, and it’s the most useful thing about me.
The Long Way Around
I spent my twenties doing physical work that most people wouldn’t - welding structural steel, operating heavy equipment, working as a medic, running large-scale event productions. For about five years I lived out of a single carry-on bag, traveling the world and taking whatever work I could find. I wasn’t lost. I was learning how things actually work.
Somewhere in that time I noticed something: I’d always loved building and mending with my hands, but as the projects got more complex, the most interesting part shifted. A large event isn’t really a physical thing - it’s a four-dimensional object, constructed in time as much as space. Thousands of interdependencies, cascading failures, systems that have to work the first time because there is no second time. I was hooked.
When COVID-19 arrived in 2020 and the physical work dried up overnight, I saw an opening. Remote work was exploding, demand for DevOps engineers was real, and I had spent years developing the operational instincts the job required. I taught myself Kubernetes, got my foot in the door, and discovered that cloud infrastructure is just another kind of building - more abstract, but governed by the same underlying principles. Systems have dependencies. Failures cascade. Design matters from the beginning because retrofitting is always more expensive. There’s always more than one right way to build something, and new ways are always evolving to meet the needs of a changing world.
In less than 2 years I was a Senior Cloud Architect, because I think in patterns and infrastructure, and I build compulsively.
What I Actually Work On
I care about tools that give people power they wouldn’t otherwise have.
Stringboard exists because investigative journalists are increasingly required to be data scientists just to find the story buried in a leaked dataset. That’s a barrier that serves the powerful and undermines accountability. I built a platform that consolidates the full data journalism workflow - acquisition, processing, analysis, secure communications - into something accessible to reporters who didn’t get a CS degree. The AI agent architecture I’m developing for it solves a problem nobody else has quite cracked: how do you give journalists the analytical power of a large language model without exposing their sources to cloud infrastructure that can be subpoenaed?
Open Hardware Manager exists because I watched COVID-19 expose something that should have been obvious: we have distributed manufacturing capacity everywhere, but no coordination infrastructure to activate it. Over 1,800 maker organizations produced 48 million medical supply units during the pandemic. Around seventy percent of our efforts went to manually solving coordination problems rather than production or design. I co-developed the data standards that make distributed manufacturing legible to machines, then spent three years building the reference implementation that makes those standards actually useful. It isn’t widely adopted yet. I’m working on that. I think it matters enough to keep going regardless.
Production AI systems are what I do professionally - agentic systems for enterprise customers across healthcare, financial services, and SaaS. I care about building things that work in the real world under real conditions, and I build them with security and safety foremost in mind.
How I Think
I can’t help noticing connections across domains that are usually kept separate. Emergency medicine taught me that most failures are systems failures, not individual failures - and that under pressure, the only thing that saves you is having built good systems before the pressure arrived. Event production taught me that complexity is manageable if you understand dependencies. Welding taught me that the joint is only as strong as the preparation, and that you can’t hide bad work indefinitely.
These things are all the same lesson. I apply it to cloud architecture, to AI systems, to open source standards development, and to the longer arc of what technology can actually do for people if it’s pointed in the right direction.
I believe we already have most of the tools we need to solve most of the problems that matter. The gap is rarely technical. It’s usually about whether the solution can be implemented in a way that concentrates wealth and power, or whether it distributes them. Open source infrastructure, privacy-preserving tools, and distributed manufacturing systems are interesting to me partly for technical reasons and partly because they represent a different answer to that question.
On Self-Promotion
I am constitutionally bad at it. I am told this is something to work on.
What I can tell you is that I do the work regardless of whether anyone is watching, I write documentation because I think other people’s time matters, and I build things I believe in even when there’s no business case for doing so. If that sounds like someone you’d want to work with, my contact information is below.
Currently: Senior Cloud Architect & Forward Deployed Engineer at DoIT International
Location: Tacoma, WA
Open to: Conversations about hard problems, collaboration on open source infrastructure, and roles where technical depth and values alignment both matter