about
I'm a first-year at Cornell studying Computer Science and Operations Research, and one of 45 Rawlings Presidential Research Scholars, which mostly means I get to start doing research early.
Right now I'm a data scientist at the Max Planck Florida Institute for Neuroscience. I showed up as an intern and ended up a co-author on a paper about how flies walk.
Before college I started OpenImg, a nonprofit that screens for genetic conditions from a single photo, and captained FIRST Robotics Team 179 to the top 16 in the world.
These days I've fallen down the markets rabbit hole: options, order books, and prediction markets. I'm looking for a summer 2027 internship in ML, or anything quant.
experience
-
Max Planck Florida Institute for Neuroscience
Data scientist
2025 – now, started as an intern
Helped figure out which neurons make a fly walk the way it does. I built the tool we used to fix limb-tracking mistakes and ran the PCA on leg movement that ended up in the paper.
- 50% fewer tracking errors
- 3× faster than planned
- bioRxiv co-author
-
FRCCast
Founder, built the whole thing
2026 – now
Live match predictions for FIRST Robotics, plus a play-money prediction market so people can bet against the model. Real-time data from The Blue Alliance and Statbotics.
- 800+ users at launch
-
OpenImg
Founder
2024 – 2026, nonprofit
Deep-learning models that screen for serious genetic conditions from one photo of someone's face, built and checked with a licensed physician.
- 98% Marfan detection
- 80%+ diabetes screening
-
FIRST Robotics Team 179
Captain + autonomy lead
2022 – 2026
Wrote the autonomous code for three straight World Championship runs and built vision that spots game pieces in real time. Also wrangled 30+ students across mechanical, electrical, and software.
- Top 16 in the world, 2025
- 1st Orlando + South Florida, 2023
projects
Stuff I built because I wanted to know the answer.
-
Volatility surface engine
Option pricing in C++ with a volatility surface that can't be arbitraged, plus why splines fall apart on real quotes.
-
Hedging Uniswap liquidity
Treats a Uniswap v3 position like an options trade, replays 2.2M real swaps to the wei, and hedges the losses.
-
Belief-state circuits
Do small transformers secretly learn Bayesian filtering? Pretty much, and I went looking for where it happens inside.
-
Micro-price on Kalshi
A classic fair-price formula breaks when prices are stuck between 0 and 1. I found where, and a fix that cuts the error 62–72%.
-
Predicting
random.randint()A huge neural net gets 54.6%. One linear layer gets 100%. Turns out Python's RNG is just linear algebra.
-
Pixel art, actually
15.8% of the biggest "pixel art" dataset isn't pixel art. I checked first, then trained a model that draws at native resolution.
-
FacePred
Guesses the word you're about to say from tiny face movements, before you actually say it.
-
Reach
Control your Mac from across the room with AirPods head tracking, your eyes, and hand gestures.
say hi
Looking for a summer 2027 internship in ML, or anything quant. Or if you just want to talk markets, my inbox is open.
JasonPitchford@gmail.com