I build bespoke AI automation systems for multi-location operations. End-to-end. The infrastructure that makes an AI system actually run on a Tuesday at 3am when nobody is watching.
I'm an AI Automation & Full-Stack Engineer. I run a one-person studio called Formaum. I work with multi-location operations across Canada, the US, and the UK that have outgrown the stack they started with and need someone who can audit, architect, build, deploy, and optimise the next one.
My work is bespoke, not templated. I pick the right tool for the job, not the one I happen to know best. Most of what I ship runs on Claude (the underlying engine), GoHighLevel or a similar consolidated CRM, and a custom layer of AI agents and workflows that sit between the tools and do the work humans were doing manually.
I came up through two industries that taught me how operations actually work at scale.
Film and TV. Production management on $100M+ productions including Skyfall, Avengers, and Game of Thrones. Sprints. Definition of done. Coordinating fifteen vendors who all need to ship the same shot on the same day. The schedule is the product. (IMDB →)
Gaming. EA Sports FIFA. A $7B global franchise with annual release cycles, multi-region rollouts, and operational complexity that doesn't forgive sloppy systems. I learned what real production management looks like at scale.
Both of those industries had the same hidden role: a human manually moving data between five disconnected tools. I was that human. I knew it shouldn't be like that, and I knew the fix wasn't a new tool. It was a real system.
So I built one. Then I left to build them for other people.
Operations leaders at multi-location businesses with real revenue and real operational drag. Industry doesn't matter. Education, franchises, professional services, real estate, gaming, entertainment. What matters is whether the operational complexity is real and whether the buyer is senior enough to make a decision and stick to it.
Recent work has covered:
If you have a sense that your operations are held together by people instead of systems, you're probably right, and we can probably fix it. (See the work →)
Production methodology. Sprints. A clear definition of done before anything ships. No surprises. The standard 6-week engagement is structured as audit, architect, build, deploy, and a final week of optimisation and handover. After that, most clients move into a monthly retainer where I keep the system improving.
I do the work. I'm not a strategist who hands off to a team in another country. The person you talk to on the discovery call is the person writing the migration scripts at 11pm on a Wednesday. That's the entire pitch.
The thread: Every client gets the same standard. Audit. Architect. Build. Deploy. Optimise. The shape of the work is repeatable. The build itself is bespoke.
Claude (Anthropic) for the AI layer. Claude Code for the build layer. Trigger.dev and n8n for orchestration. Supabase, PostgreSQL, BigQuery for data. GoHighLevel, HubSpot, Salesforce, Monday for CRM (depending on what's right for the operation). Twilio, Vapi, ElevenLabs for voice. Next.js, FastAPI, Python for custom apps.
Tools change. The methodology doesn't.
I'm based out of Canada. I read, I garden, and I think a lot about how operations should be designed when the AI layer can do the boring parts. Most of my best ideas come from watching someone else's team do something manually that a system should already be handling.
If that's you, let's talk.
45 minutes. No pitch. By the end of it you'll know what to fix first and whether I'm the right person to help you fix it.
Book a Discovery Call