I'm William Rex — I build real things with AI.
Fort Collins, CO · open to relocating
View my work ↓
A bit about me
I build practical tools with AI — the kind that quietly take repetitive work off people's plates.
Most of what I've made has been for a small remodeling business: systems that handle receipts and generate proposals automatically, turning hours of manual paperwork into something that runs on its own. I scope it, build it, and document it so it keeps working after I hand it off.
I like understanding a problem before building, and I'm always learning — usually by taking on something I haven't done before.
Stuff I've built
AI receipt system
Reads emailed or photographed receipts, codes them against a contracting company's cost-code system, and files them automatically — replacing a manual, paper-based process. Scoped, built, and documented so the team can run it without me.
Proposal generator
A self-serve tool that turns a few spreadsheet fields into a polished, on-brand PDF proposal in about 30 seconds — collapsing an hour of manual document-building into one click. Built as a chained pipeline (Apps Script → automation → PDF generation) and handed off so the team runs it themselves.
Claude for a real-estate broker
Worked one-on-one with a local real-estate broker: set up Claude alongside the email and Google tools she already used, then trained her to put it to work in her actual day-to-day. The focus was adoption — making it something she'd genuinely use, not just installing software and walking away.
Building & learning right now
Going deeper on Claude and building something new most weeks — across automation, agents, and generative media.
I'm curious about all of AI
Not locked into one niche — I want to understand the whole landscape, not just one corner of it.
AI agents & automation
Getting software to carry out real work, end to end.
Generative image & video
Exploring new tools like Higgsfield and Veo as they come out.
AI for real-world impact
Putting these tools to work for organizations that need them.
How the models actually work
Prompting, evaluating outputs, and understanding what's under the hood.