A complete application in a minute
Tailored resume and cover letter, ready to send.
Valen Labs builds an autonomous biofactory that designs, builds, and characterizes molecules at scale. We partner with frontier AI labs to train biology models and generate high-fidelity data for ML benchmarking.
As a Member of Technical Staff on AI Research, you design evals and benchmarks, run RL and post-training loops, and decide what data is worth generating in scale. You sit between the lab and the model, choosing experiments, signaling data needs, and proving improvements.
The next great drugs won’t be discovered. They’ll be designed. But AI can’t design biology it has never seen, and the data to teach it barely exists yet.
Valen Labs is building the engine that generates it. We are a biological data company. We’ve orchestrated novel technology and ambitious lab machinery into an autonomous biofactory that designs, builds, and characterizes molecules at a scale and speed the field has never had. Best in class machines, best in class data.
We work with frontier AI labs to train their biology models, and we build the benchmarks that measure how well AI can actually do science, from a single binding assay to running an entire lab. Everything we produce is structured, high-fidelity data, built for machine learning and generated faster than anywhere else.
We’re backed by top Silicon Valley investors, including funds linked with Peter Thiel, and we’re based in San Francisco.
Valen’s data only matters if it teaches models something true. That is your job.
As a Member of Technical Staff on AI Research, you design the evals and benchmarks that measure what AI can actually do in biology, you run the RL and post-training loops that turn our data into better models, and you decide what good data looks like before we generate it at scale. You sit between the lab and the model: you choose which experiments are worth running, what signal they need to carry, and how we prove a model got better. The benchmarks you build become the yardstick the field uses to measure AI in biology.
What you’ll own:
Design and build evaluations and benchmarks for AI across the spectrum of biology, from single-assay prediction to running a full experimental campaign
Run RL and post-training experiments that turn Valen’s data into measurably better biology models
Set the standard for what good data looks like, and tell the bench team what to generate and why
Work directly with frontier AI labs on how our data trains their models
Close the loop between model performance and experimental design, so every run makes the next model better
Hands-on experience building evals or benchmarks for ML systems, especially for models whose outputs are hard to score
Direct experience with RL, post-training, or related model-improvement work
A real sense of what good data looks like, and the judgment to tell signal from noise before anything is labeled
Strong engineering: you can design the experiment, build the pipeline, and ship the analysis yourself
You don’t need a biology background, but you need the hunger to learn it fast
A bias toward speed and ownership on a small team
San Francisco based, fully onsite
Medical, dental, and vision coverage
Meaningful early-stage equity
Unlimited PTO
We work hard, and we’re honest about that. The pace is high and the problems are hard. What you get in return is work that genuinely matters, real ownership, and a small team that holds a high bar and ships fast. This is a place to do the best work of your career.