Staff Machine Learning Engineer
Summary
Staff ML research engineer (placed by Fruition Group at an autonomous vehicle company) building large vision-language-action driving models: shaping data, running large training runs with SFT and RL post-training, evaluating on real vehicles, and compressing models to run onboard. Key stack: PyTorch/TensorFlow/JAX, transformers, diffusion, ONNX/TensorRT. On-site SF Bay Area; $200k-$260k base + equ
Staff Machine Learning Research Engineer, Multimodal Driving Models
Location: San Francisco Bay Area, CA (on-site)
$200,000 - $260,000 base + equity
About the Role
Fruition Group is exclusively partnering with a fast-growing autonomous vehicle company to find a Staff Machine Learning Research Engineer. Our client is building large multimodal models that look at what the vehicle sees and work out what it should do next, and they want strong researchers who can take those models from experiments to running on real vehicles.
This role suits someone who enjoys the full journey: shaping the data, designing and training the model, proving it works in realistic driving conditions, and then making it small and fast enough to run onboard. You’d be working on a problem the industry is only just starting to crack, with the backing and the vehicles to test your ideas.
What You’ll Do
- Build, train, and test vision-language-action models that propose high-level driving choices and guidance for the vehicle’s path.
- Take responsibility for a major research thread, covering the data, the model design, large training runs, and checking results on actual vehicles.
- Create training and evaluation pipelines, along with demanding measures of how well models perform in driving situations.
- Work out ways to shrink large models so they can run on the vehicle’s own computer without losing what makes them useful.
- Use supervised fine-tuning and reinforcement learning after pre-training to improve reasoning, reliability, and handling of unusual situations.
- Work closely with the perception, planning, and platform teams to carry successful research into the production system.
What Our Client Is Looking For
- A master’s degree in computer science, electrical engineering, mathematics, statistics, or a related field; a PhD is a strong advantage.
- 5+ years building and training deep learning models in PyTorch, TensorFlow, or JAX.
- Hands-on experience training vision-language or vision-language-action models.
- A record of taking models from training and evaluation through to production use.
- A deep grasp of the current landscape: multimodal models, transformers, diffusion, and flow matching.
- Experience with large-scale or distributed training.
- Model compression and inference optimization (quantization, ONNX/TensorRT, mixed precision, custom kernels), post-training of large multimodal models, work with camera, lidar, and radar data, publications at leading ML and robotics conferences, or a background in autonomous driving or driver assistance.
Apply Now
Ready to take cutting-edge driving models from research to the road? Apply now by sending your resume to Levi Barnes at Fruition Group - levi.barnes@fruitiongroup.com. All inquiries are handled in confidence, and applications are reviewed on a rolling basis, so don’t wait to reach out.
Skills
Sign in to read and join the discussion.