Implementation, Control,
and Practice
Production software, delivered fast — with support that runs on AI, not headcount.
Book Appointment
30+ Years Tech Experience. Two modes: I build it for you, or I work alongside you until you can build it yourself. Either way, the engagement ends with something real — working software, a fixed system, or a team that knows what it's doing.
I don't just show up to write code — I get in and understand how the business actually operates first. When it's called for, that means process re-engineering, not just automation: finding where a workflow is actually broken before deciding whether AI, custom software, or neither is the right fix.
Six of those years were on Wall Street, managing infrastructure, technical support operations, and call center ops — the part of the job where systems break and someone has to own the outcome. That background shapes how I approach AI work: I design for what happens when things go wrong, not just when they go right.
That infrastructure instinct goes back further — I ran a computer configuration center in the 90s, building hardware and installing software by hand, and watched that whole discipline turn into automated provisioning in real time. I was doing automated builds back then — systems provisioned automatically to a specific user's spec, not just imaged off a master disk. Same problem I solve today with cloud instances and Ansible; the tools changed, the discipline didn't.
After that, I ran my own bookkeeping company, building and running QuickBooks integrations for clients — hands-on with the unglamorous plumbing that actually keeps a business's numbers straight, not just the theory of it.
I've also built for a federal government contractor: an ERP connector that took a week of manual work down to about a minute of input and ten minutes of waiting. That's the kind of leverage I'm looking for on every project — not a demo, an actual week given back.
My focus is control: how models behave in real environments, where workflows break, and how to design systems that keep humans in charge of outcomes.
This is not tool evangelism. It's practical leverage — architecture, implementation, and hard-won judgment — using the models available today.
Build & Deliver
I build it. You ship it. Bring a project, a broken workflow, or a problem you haven't been able to solve. I architect and build the solution — production-grade, not a demo. Small-company budget, big-company feature set.
- Web apps, APIs, and backend systems
- AI integrations into existing products or workflows
- Autonomous post-launch support, powered by Fire — not a ticket queue you have to staff
- Correcting unstable or poorly designed AI implementations
- Architecture reviews with concrete output
- MVP development from concept to working software
Transfer & Uplift
I build it with you. We work on your project together. You leave with a working system and the knowledge to maintain, extend, and build the next one yourself.
- Hands-on AI development workflow from scratch
- Building and controlling agents in real environments
- Understanding failure modes before they hit production
- Translating intent into reliable, repeatable system behavior
- Upskilling a technical lead or small team
Support That Doesn't Need You
Every build ships with a support layer that runs on Fire — the same model-agnostic inference engine already running production traffic across five live products in the Modus suite. Your customers get answers, your system gets watched, and you don't make a first support hire to make it happen. It isn't a pitch — it's the same engine, pointed at your product.