Basketball
UConn, specifically, and to a degree that mildly concerns the people around me. I know more about the program’s history than is strictly useful.
Tristan Couvares
I find the value in people and systems before it’s priced in — then I build.
I’ve helped take companies from a dozen people to a few hundred, been close enough to a billion-dollar IPO to feel it, and I still write code most weeks to fix things that annoy me.
A recurring operating question, made testable.
Hiring
Capacity
Lab
An interactive, synthetic model for exploring hiring demand, recruiting throughput, and backlog assumptions.
A weekend project before the feature was everywhere.
GPT in
iMessage
Wired a language model into iMessage, back when that was a weekend project rather than a product in everything.
Boring. Finicky. Used constantly.
CSV Import
Reshaper
Reshapes messy CSV exports into the exact form a finicky destination will actually accept.
Because backtracking is quietly satisfying.
Sudoku
Backtracking
Solver
A backtracking solver, written mostly because backtracking is quietly satisfying to watch work.
The receipts
Bets that
paid.
- being early, for a living
- 15+ yrs
- people, from a dozen up
- ~250
- TrueCar IPO, close enough to feel it
- $1B
- dual citizenship
- US + EU
- Public-company scaleEarly at TrueCar
Close enough to watch an early bet grow into a billion-dollar public company.
- Team scale · ~12→250Electric-vehicle startup
Helped the organization grow from around a dozen people to roughly 250.
- Independent · 10 yearsUnlikely matches
Ran a small operation connecting people with companies most had not heard of yet.
- Current workAI products
Working on the people and operating systems behind a company building AI products.
The titles vary. The job is always the same: be right about people early.
In public
Arguments worth keeping.
-
Hiring systems
Just Say No to Assessment-First Hiring
A practical boundary for take-homes, one-way interviews, and other hiring assessments.
Read → -
Markets
Caveat Emptor
On interdependence, venture funding, and what happens when the music stops.
Read → -
Markets
Why you shouldn’t listen to VCs
A contemporaneous argument against the loudest version of the 2016 bubble chorus.
Read →
About
The short
version.
I’ve spent fifteen-odd years doing essentially one thing under a rotating set of titles: being early on people and companies, when the evidence is thin and the price is low. The seat changes — the instinct is older than any of the jobs. I was a kid taking machines apart, a teenager posting code to forums, a sysadmin before I could legally drink.
I think about the world in terms of edges and asymmetries: where is something mispriced, what’s the downside if I’m wrong, what’s the upside if I’m right. That frame works on hiring, on markets, on code, on nearly everything I find interesting. I’d rather be dangerous across a few domains than narrow in exactly one.
Currently
What I’m working on now
I’m working on the people and operating systems behind AI products, taking algorithms coursework for fun, and turning recurring recruiting problems into small tools that can be tested in public.
Updated August 2026
Otherwise
What occupies me the rest of the time
Thin markets
How scarcity gets priced when almost nobody is trading. I collect a few things, mostly as an excuse to keep thinking about why they’re worth what they’re worth.
Uncertainty
Priors, base rates, and the gap between a good decision and a good outcome. Most people confuse the two; I try not to.
Staying dangerous
I take real CS coursework for fun — an algorithms sequence out of Stanford, a Python specialization — because I’d rather understand the machine than rent it.
Your move
Say
hello.
I’m easy to find and I read everything, even when I’m slow to reply. I’m especially interested in AI operating problems, talent systems, and unusually early bets — though arguments about basketball are always welcome.
hello@tristancouvares.com