Job Description

We're building AI capable of autonomous scientific discovery. Our mission is to reach a world where discovery is no longer bottlenecked by the number of human scientists alive to pursue it. Reinforcement learning has taken models remarkably far in math and code, where an answer can be checked and a model can learn from ten thousand attempts. Science offers no such reward function. The judgment a scientist exercises before any result is certain, which hypothesis is worth the month, what the evidence actually supports, when to abandon a promising dead end, has never been captured at scale. That missing signal is the real bottleneck, and it's the one we build for: the environments, datasets, and expert reasoning that teach models to do science, not just recite it. That work is already in the world. We built and operate Aristotle, an AI research partner trusted by thousands of researchers at some of the most demanding institutions in science, including GSK, Pfizer, AstraZeneca, Merck, Stanford, MIT, Harvard, and the FDA. Alongside it, we're developing autonomous research systems that take on long-horizon scientific problems end to end. Backed by leading investors and advised by the scientists who built the field, Autopoiesis is redefining what discovery looks like in the most important industry of the century. About the role We're looking for a Member of Technical Staff to work on the core problem of training models that can do science on their own. This is a research and engineering role in equal measure. You'll run experiments, build the systems that make those experiments possible, and turn what works into models that push past what's currently possible. We don't draw a hard line between research and engineering, and the best people here don't either. Designing a new environment, shaping a pretraining run, and getting more out of the training stack are all part of the same job, and doing them well means moving between the idea and the implementation without losing the thread. 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. You'll have real resources behind you, including access to significant compute, the freedom to run the experiments worth running, and a team that will help you get sharper fast. We care more about how you think and what you've built than about credentials or years in the field. Key responsibilities Drive research on autonomous scientific discovery, from pretraining to the environments and reward signals that teach models to reason like scientists. Design and run experiments end to end, forming hypotheses, building what tests them, and following the evidence. Build and improve the training stack, making runs faster, more reliable, and more ambitious. Turn research into real capability, closing the gap between a promising result and a model that works. Work closely with researchers and the founding team on the hardest parts of the core problem. Preferred qualifications Experience with pretraining, large-scale training, or reinforcement learning. A track record of research contributions, shipped models, or systems others relied on. Familiarity with scientific domains or a genuine curiosity about them. Early-stage startup experience, or the temperament for it. Our values The mission is the only thing. Everything we do either moves discovery forward or it doesn't, and anything that doesn't is a day we won't get back. We hold every decision, every hire, and every week against that one question. The mission is hard enough that nothing else earns a place next to it. Bet on the process, not the outcome. We think in expected value. A good decision made on the best available information is a good decision even when the result disappoints, and a lucky win off a bad process teaches us nothing. We reward sound reasoning, run the bets worth running, and judge ourselves by how well we chose, not only by how it landed. Don't compete, be uncontested. Competition is a symptom of a weak position. We don't win by racing others down a crowded path, we win by building something no one else can, in a direction no one else is looking. When we find ourselves fighting for scraps, we've made a strategic mistake upstream. Own it end to end. There are no bystanders here. We spot what's broken, take full responsibility for a strong result, and see it through. We're self-driven, own our mistakes plainly, and feel deep responsibility for what we're building. Where we work We are hiring for this position in our San Francisco office. We are in person 5 days a week. What we offer This role offers competitive salary, meaningful equity, and competitive benefits, along with the compute and support to do the most ambitious work of your career. You'll have unusual freedom to chase hard problems, which also means we'll expect a lot. Equal opportunity Autopoiesis Sciences, Inc. is an equal opportunity employer committed to diversity and inclusion. We welcome applications from all qualified candidates regardless of race, gender, age, religion, sexual orientation, or any other legally protected characteristics. #J-18808-Ljbffr

Job Tags

Similar Jobs