Hey — I'm Liam. I build web products and AI systems, and I've been doing it long enough to have collected some scars, learned some expensive lessons, and developed the calm energy of someone who's debugged a production incident at 2 a.m. with nothing but logs and a prayer. I didn't take the straight path here. I studied anthropology (turns out understanding humans is wildly useful in engineering), spent years writing music (timing + structure + flow = underrated engineering skills), and eventually discovered I really like building things people actually touch. My work lives somewhere between engineering, product, and "sure, let's try it and see what breaks." I like scrappy teams, weird ideas, fast loops, and projects with a real heartbeat. If a project has ambiguity, moving parts, or needs someone to shepherd it from idea → prototype → reality, that's usually where I shine.
WhatIDo(andenjoydoing)
Full-Stack Engineering
TypeScript, Python, Next.js, React, Flask — clean, readable code that doesn't fight future-you.
AI / RAG / Multi-Agent Systems
I build practical AI features that actually help people. Not "look what the model can do," but "look how much faster this makes your workflow."
Zero-to-One Product Building
Give me a vague idea and a couple constraints — I'll turn it into a thin slice we can put in front of users by next week.
Turning Chaos Into a Plan
I'm good at translating: founder vision → engineering tasks → something we can ship without sprinting off a cliff.
SelectedWork
Citywide Dining Week App
A nonprofit needed a full event app. I designed and built the whole thing with Next.js + Supabase. Result? • 3,557 users the first week • No frantic Slack messages • Volunteers ran the event from their phones like it was nothing Simple, clean, durable. My favorite kind of build.
AI Automations for a Performing Arts Org
Transformed a pile of tribal knowledge into AI workflows that handle audio analysis, visual classification, and documentation. It made their entire process… less magical, more replicable. (That's secretly the whole point of AI.)
Dark Web Monitoring Toolkit
OngoingA lightweight threat-intel app with ingestion, API endpoints, telemetry, and dashboards — built intentionally without the usual security-tool bloat.
RAG + Agent Systems for Startups
Retrieval pipelines, guardrails, model orchestration, structured outputs, observability — in English: I make "AI features" behave like grown-up software.