Research · Santa Clara, California (on-site) · Full-time
Research engineer, foundation models
Train AREN. Own a piece of the foundation model stack end to end, from data curation through evaluation.
title: Research engineer, foundation models location: Santa Clara, California (on-site) team: Research employment: Full-time posted: 2026-04-01 summary: Train AREN. Own a piece of the foundation model stack end to end, from data curation through evaluation.
About the role
You will be a core contributor on the AREN foundation model. Expect to own a piece of the stack end to end — data curation, training infrastructure, evaluation, and the weekly argument about what the model should learn next.
This is not a specialist role. We are looking for generalists who can reason about architecture, datasets, and systems trade-offs in the same conversation.
What you would work on
- Architecture and training code for multimodal transformers over vision, LiDAR, radar, and CAN streams
- Data curation pipelines, including simulation-authored curricula
- Evaluation harnesses that reflect how the model will actually be used in a vehicle, not synthetic leaderboards
- Ablations that change what the rest of the team believes about the model
What we expect
- Five or more years of applied machine learning experience, with meaningful time on model training at scale
- A habit of reading papers critically and designing experiments that actually resolve disagreements
- Comfort with PyTorch, distributed training, and reading CUDA when it matters
- A low tolerance for unverified claims in your own work, and in others'
What we do not require
A PhD. Publications. A specific institutional background.
Logistics
- On-site in Santa Clara, California. We do the work together, in person.
- Full-time, full benefits
- Compensation is competitive with frontier AI labs. Exact bands on request.