In short
San Francisco, CA (South Park office, in person 5 days per week) Compensation: $200,000 base + highly competitive early-stage equity Employment Type: Full-time Visa Sponsorshipโฆ
PYTHONPYTORCHLLMMACHINE LEARNING
๐ About the job
๐
LocationSan Francisco, CA (South Park office, in person 5 days per week)
๐ฐ
Pay$200,000 base + highly competitive early-stage equity
๐๏ธ
Job typeFull-time
Machine Learning Researcher
Company: Confidential โ venture-backed technology company
Visa Sponsorship: Visa transfers; can sponsor visas
About the Company
The client is pursuing general causal intelligence: AI that can predict the future and identify the actions that change it. It is building a Large Physics foundation Model (LPM), because domains governed by physics have inherent cause-and-effect structure that visual or textual data lacks. Its starting domain is weather, the most observed physical system on earth, with rapid ground-truth feedback and data volumes that dwarf LLM training sets.
The founders come from Cruise, Google Research and Meta. The company is about 10 people in San Francisco, growing to around 35 this year, and is backed by Kindred Ventures, Refactor and BoxGroup.
The Role
The client is hiring ML researchers to build powerful physics models grounded in observable feedback and verifiable ground truth. If you have done frontier research and trained large-scale models from scratch in language, vision, robotics or biology, you will work across the full ML stack to build a Large Physics foundation Model, starting with weather.
What You Will Do
๐นWork across the full ML stack: data, model, evaluation and infrastructure.
๐นImplement novel model architectures and training algorithms.
๐นBuild data pipelines and training infrastructure for massive, petabyte-scale, multimodal datasets.
๐นIterate rapidly on experiments and ablations, and read results through careful analysis.
๐นBring new ideas from current research into the work.
What You Bring
๐น1+ years training large-scale foundation models from scratch (pre-training ideal), not only fine-tuning
๐นStrong ML fundamentals and depth in at least one core domain (computer vision, sensor fusion, language models, physics-informed networks)
๐นDistributed training and inference across hundreds to thousands of GPUs
๐นPetabyte-scale data pipelines for model training
๐นA mission-driven passion for science or the physical world
๐นIndividual-contributor focus
๐นAbility to work in person in San Francisco 5 days a week
Nice to Have
๐นResearch lab experience (e.g. DeepMind, FAIR)
๐นPhD or Master's in a quantitative field; background in physics or related sciences
๐นMeteorology, computational fluid dynamics or numeโฆ