One experiment
a lab would hire for.
In 30 days, derive, run and write up one experiment on a real model, the work applied-AI and research teams at frontier labs do.
Not sure yet? Am I ready for more?
You already know what the loop will ask
How did you measure it? How noisy is that number?
What would have proved you wrong?
The cohort gives you one piece of work that answers those questions.
The month
Derive it
Pick one question a lab would care about and write down the math that predicts the answer, before any code runs.
Run it
Build the experiment as a Colab notebook on Python 3.12: a baseline, a change, and a measurement on a real model, with error bars.
Write it up
Turn the result into a write-up a research team would accept, and practice defending it the way an interview loop will ask you to.
Live sessions each week, plus a library you keep. The cohort is small enough that your experiment is the worked example.
Built for you if
- You are preparing for an applied-AI or research role at a frontier lab
- You write Python, or can follow a derivation and will learn fast
- You can give the month several hours a week
- You want a piece of work you can show
Not this time
- You want a guaranteed offer or a referral
- You want the answers without running the code
- You have not picked a direction yet: start with the free book
Taught by Rod Rivera
Professor of the Practice at ITAM, with a PhD and more than twenty peer-reviewed machine-learning publications. Nebius Academy lecturer. Fifteen years of industrial machine learning.
Apply for the next cohort
Rod reads every application. Tell him the role you are aiming for.
Not sure yet? Am I ready for more? · Read the material first.
