Most teams are bolting AI onto the way they already work. I think that’s backwards.
I’m a lean and agile minded engineering leader with 17 years in software covering web, mobile and mlops across in industries spanning security, fraud prevention, ecommerce, telecommunication, and recently in banking. Most of the job now is keeping people, code, and agents happy.
Evenings and weekends I build small systems that test where that balance actually lands.
All of it runs in production. I’d rather argue with something you can click on.
The work
-
Daily Kickoff
A news feed for people building with AI.
Up to seven stories a day, picked by an LLM pipeline that runs itself every night. No cookies, no tracking, no ads, and no JavaScript on the page.
-
Mental Models Explorer
Semantic search across the mental models people actually use to decide.
159 models from 13 thinkers, searchable by meaning rather than by keyword, with a graph of how they connect. The whole thing sits on a free tier.
-
Wealth Forecaster
Will the money last?
A lifetime balance sheet: the saving years, then spending it back down, with inflation, tax, a pension and one-off costs. It runs in your browser and nothing gets sent anywhere.
-
Points Optimizer
What are your credit card points actually worth?
It starts from what you want to book rather than what the points are nominally worth. Works offline and installs as an app.
-
ContractIQ
Semantic search over City of Richmond procurement contracts.
Built with a team at Hack for RVA, across 1,375 public contracts. The repo belongs to the team rather than to me.