AI for future doctors, researchers & public-health innovators — train a real model and an AI research crew, responsibly. Not an “AI doctor.”
This is education — not medical advice. Students use public, built-in teaching datasets — never real patient data. The lab does not diagnose, treat, or advise anyone, and AI is always checked by a human. A core goal is learning where medical AI fails. Real decisions require licensed professionals.
Students combine real medical machine learning with the AI-agent skills from Level 1 to build tools that analyze and explain — while thinking critically about every result, the way a good clinician does.
Train a classifier on a real medical dataset — and learn why a missed case (false negative) is the most dangerous error.
Four agents — Researcher, Data Analyst, Ethics & Safety, Writer — that explain research and flag its limits (built with CrewAI).
Data Explorer, a prediction model with a live threshold slider, and a research-brief generator.
Four weekly 2-hour live sessions. Small groups, with all datasets and AI access provided — nothing to install or pay for.
It's easy to build a model that looks impressive. The real skill — the one medical schools and mentors respect — is understanding its limits. Students leave able to question an AI result like a clinician, not trust it blindly.
A real web app students build and demo — with a live threshold slider as the star feature.
Dataset overview, features, label, charts, and the data card.
Accuracy, confusion matrix, sensitivity/specificity — and a live threshold slider.
Paste an abstract → a safe, structured one-page brief with limits.
8 hours of live instruction over 4 weeks
No. Absolutely not. This is fully educational, using public, built-in teaching datasets. The lab does not diagnose, treat, or advise anyone, and uses no real patient data. The goal is to teach how medical AI works and where it fails.
It's strongly recommended — this Level 2 lab builds on the AI-agent skills from Level 1. Students with solid Python and AI basics from elsewhere can also fit; reach out and we'll help. Enrolled students get Python and AI-math prework on our portal.
Both, by design. Students build genuine AI tools used in healthcare research, but the heart of the lab is thinking critically about results — bias, errors, ethics — the way a good clinician does. That combination stands out on medical-track applications.
Only public, well-known teaching datasets that are built into the tools (no downloads, no real patient data). We use them to teach the science and the ethics — including how datasets can be biased.
A working AI Medical Research dashboard, a one-page Health Research Brief for their portfolio, and a Research Ignited Level 2 certificate. Strong students can extend the project into our Research Program.
A standout project, an honest education in AI's role in medicine, and a credential that means something. Seats are limited.
Reserve a Seat →