Software Engineer, ML Infrastructure

About Arc Institute Arc Institute is an independent nonprofit research organization at the interface of artificial intelligence and biology, working to accelerate scientific progress and understand the root causes of complex diseases. Founded in 2021 and based in Palo Alto, Arc partners with Stanford University, UC Berkeley, and UC San Francisco. Unlike academia, our scientists have long-term funding and industry-like resources. Unlike industry, they're free to pursue high-risk, long-term research without commercial pressures.

Arc's Technology Centers and Core Investigator labs work side by side, integrating experimental and computational biology under one roof to tackle problems neither could solve alone. Our two Institute Initiatives reflect this model in action: Virtual Cell Initiative : Building a full-stack virtual cell model to identify disease mechanisms and nominate drug targets, accelerating the path from biological insight to clinical trials. Alzheimer's Disease Initiative : Mapping the genes, pathways, and environmental factors behind Alzheimer's disease to develop drug candidates that address root causes.

More than 300 Arconauts work together at our Palo Alto headquarters, backed by substantial long-term philanthropic funding. About the position You'll join Arc's Infrastructure Engineering team and own the data and evaluation of ML infrastructure for Arc's virtual cell models. You'll start with two systems: our Model Performance Dashboard, which benchmarks those models, and the pipeline that turns single-cell data into their training sets. You'll set the technical direction for both and connect them, so every new dataset version is judged by what it does to the models.

Later in your first year, the role widens into the engineering that makes the ML team's training and evaluation run fast and at scale, lab-in-the-loop infrastructure, validation frameworks for agent-written software, and Arc's rules for what research data we bring in, keep and share. The role consolidates ownership that is split across several teams today. Why this position could be the best job you’ve ever had You will contribute to building a virtual cell. Your evals show which data makes Arc's models better, and the answer shapes which experiments Arc's labs run next. You will own the platform end to end, and you work directly with the ML researchers and scientists who rely on it.

About you You finish what you own, and you fix what's broken without being asked. You're curious about the biology, enough to learn single-cell perturbation data and argue about what an eval should measure. You're steady when priorities shift. Research is high-risk, so you test ideas early and change course when the evidence says to. You don't wait on scarce expert time. If proving a dataset's value means running the model evals yourself, you run them. In this position you will Own how Arc measures the performance of its virtual cell models, starting with the backend of our Model…

Read the full posting