keanan.w
← Work
Title
Mathcom
Role
Founder
Year
2024 – 2025
Mathcom

It's easy to build when there's pressure. The gap year taught me I build just as hard when there isn't.

15
person volunteer team
1yr 4mo
start to wind-down

The story

I'd already gotten into UBC. Gap year, nobody watching, no deadlines, no grades, no admissions officer to impress. I built a nonprofit math learning platform alone, for no external reason except wanting to see if I could.

It ran three departments: an AI and Web team shipping MathCom AI, a tutoring assistant that was deliberately not a calculator; a Content team making math feel human through reels and explainers; and a Social team running charity outreach.

The real work

The code was Flask REST APIs with streamed responses and usage logging, a dozen behavioral guardrails across two LLMs. But the hard part wasn't code. It was recruiting and managing 15 volunteers who weren't paid and weren't obligated to show up. That's a completely different problem, and it's the honest answer to how I grew that year.

I wound it down on purpose. It didn't fizzle, I ended it, and that's the part I'm most sure about.

Stack  Flask, Python, LLM integration