JOB-2145
A venture-backed Physics AI company that helps engineering teams replace slow simulation loops with machine-learning models is hiring forward deployed applied AI/ML engineers to build and ship models on real customer problems.
Your focus
- Own model development for customer engagements: architecture selection, training pipelines, tuning, evaluation and validation against simulation ground truth
- Work with applications engineers to make sure training data is physically meaningful for each use case
- Partner with data and platform engineers on training pipelines, model registries and inference serving
- Apply state-of-the-art methods such as neural operators, diffusion models and geometric deep learning to customer problems
- Co-engineer side by side with customer data scientists and bring field insights back to research and product
Ideal background
- 5+ years of hands-on applied ML on engineering or simulation data (exceptional PhD graduates with 12 years considered)
- Experience training surrogate models with Python and PyTorch on CAE data where geometry and mesh change between samples
- Hands-on with mesh, graph or point-cloud architectures for physics surrogates
- Experience deploying ML into real engineering workflows in a customer-facing or consulting role
- U.S. person status (citizen or permanent resident) for defense work
On-site in San Francisco, CA. Competitive equity.
