Internship
Autopoiesis Sciences
Summary
We're looking for exceptional interns to work alongside our research team on the core problem of training models that can do science on their own. This is a real research and engineering role, not a shadowing exercise. You'll take on genuine problems, run experiments, build the systems around them, and see your work make it into what we're building.
Job description
About the role
We're looking for exceptional interns to work alongside our research team on the core problem of training models that can do science on their own. This is a real research and engineering role, not a shadowing exercise. You'll take on genuine problems, run experiments, build the systems around them, and see your work make it into what we're building.
We don't draw a hard line between research and engineering, and we don't want interns who do either. What matters is that you can form a hypothesis, build the thing that tests it, read the results honestly, and decide what to do next.
We run internships in three cohorts, Fall 2026, Spring 2027, and Summer 2027. Each is a 12-week paid internship, in person at our San Francisco office.
What you'll bring
- Clear proof you can do hard technical work, whether research, shipped systems, or training code. Show us what you've made.
- Real depth in ML and the fundamentals that outlast any one method or framework.
- Comfort working across research and engineering, without needing either label to define you.
- Hunger, speed, and the habit of teaching yourself whatever a problem demands.
How to stand out
The best people we've met don't wait to be told what to do. If you think you're one of them, show us rather than tell us. Send us something you built, or find a way to get our attention that a standard application wouldn't. We're looking for people who stand out, and the easiest way to prove you're one of them is to act like it before anyone asks you to.
Requirements
- Clear proof of hard technical work in research, shipped systems, or training code
- Real depth in ML and the fundamentals that outlast any one method or framework
- Comfort working across research and engineering
- Hunger, speed, and the habit of teaching yourself whatever a problem demands
Benefits
- Paid Internship
- 12-Week Program
Company overview
Autopoiesis Sciences is a San Francisco-based applied AI lab founded in 2025, building artificial intelligence systems capable of conducting science autonomously. The company is the creator of Aristotle, described as the first AI research partner and "co-scientist," designed to help medical and biological researchers accelerate discoveries, validate complex hypotheses, and address urgent challenges in medicine and biology. Unlike conventional large language models that tend to affirm user hypotheses, Aristotle is engineered as a "self-skeptical" reasoning model that challenges assumptions and provides objective feedback on the likely outcomes of scientific theories. The company runs its large-scale models on Oracle Cloud Infrastructure, leveraging OCI's performance and healthcare/life sciences expertise to reduce training time and inference latency. Founded by CEO Joseph Reth, Dr. Eike Gerhardt, and former senior FDA official Lawrence Callahan, Autopoiesis aims to define the AI-for-science era and empower the world's best scientists to move faster and save lives.