10 days ago
Base Salary
$273k - $345k/yr
Responsibilities
- Research and develop reinforcement learning and distillation techniques for trajectory planning.
- Design and deploy multimodal models that translate visual perception and behavioral goals into physical vehicle actuation.
- Develop interactive world models from multi-sensor logs for trajectory re-simulation and counterfactual analysis.
- Adapt autonomous-driving models to novel environments and edge cases.
- Own model post-training, distillation, quantization, profiling, and deployment for low-latency vehicle edge hardware.
- Optimize inference pipelines and hardware-accelerated runtimes while addressing CPU, GPU, and memory bottlenecks.
- Work with hardware, electrical, and firmware teams on carrier boards, sensor interfaces, and edge GPUs.
- Architect pipelines for ingesting, filtering, and extracting rare scenarios from multi-petabyte sensor datasets.
- Build data curation, active learning, data mining, and statistical quality workflows for machine learning pipelines.
- Partner with QA, validation, testing, and simulation teams to detect regressions and convert real-world anomalies into simulation tests.
Requirements
- 10+ years of non-internship professional machine learning engineering experience.
- Deep expertise applying AI Transformers to robotics, physical actuation, or spatiotemporal data.
- Experience designing or training multimodal systems, large-scale VLA models, or generative Diffusion models.
- Strong sensor-fusion experience using camera, LiDAR, and radar data.
- Fluency in PyTorch or JAX and proficiency in Python, with familiarity or robust programming skills in C++.
- Experience with model optimization, distillation, deployment, edge hardware, profiling, or runtime compilation.
- Experience with data curation pipelines, active learning workflows, data mining architectures, or massive physical datasets.
- Familiarity with multi-task learning, Bird's-Eye-View frameworks, representation learning, spatial tokenization, robotics data structures, or low-level sensor interfaces is preferred.
- Ability to structure and derive signal from ambiguous real-world data distributions.
Benefits
- Medical, dental, vision, disability, and life insurance.
- Flexible Spending Account and Health Savings Account options.
- 401(k), equity eligibility, sick time, unlimited flexible time off, and paid holidays.
- Paid parental leave and a pre-tax commuter benefit plan.
- Team lunch in the SoMa office every Tuesday and Thursday.
- Full-time exempt role based in the San Francisco office with onsite work five days per week.
Categories
AI ResearchML Engineering
