Machine Learning Engineer — Reinforcement Learning
$150k - $450k • Sunnyvale, CA • Full-time
Posted 12d ago
Job Location
Sunnyvale, CA
Tech Stack
Remote Work Policy
On-site
Employment Type
Full-time
Categories
Machine Learning Engineer
About the job
We are a dedicated research lab focused on building, understanding, using, and risk-managing foundation models. Our mission is to advance AI research, cultivate the next generation of AI builders, and make significant contributions to a knowledge-driven economy. As part of our team, you will work on the core of cutting-edge foundation model training alongside world-class researchers, data scientists, and engineers. You will tackle fundamental and impactful challenges in AI development, contributing to groundbreaking AI solutions that have the potential to reshape industries. Strategic and innovative problem-solving skills are essential for establishing our institution as a global hub for high-performance computing in deep learning.
Responsibilities
- Extend or modify distributed training frameworks (e.g., DeepSpeed, FSDP) for new use cases and architectures.
- Integrate training workers with rollout generation engines (e.g., vLLM, SGLang), including trajectory exchange and policy-weight synchronization.
- Create and debug multi-node launch scripts with flexible batch sizes, parallelism strategies, and hardware targets.
- Build systems for experiment tracking, job monitoring, and logging for collaborators and researchers.
- Write production-quality code and tests for ML infrastructure in PyTorch or JAX, ensuring reliability and maintainability at scale.
- Implement reward/verifier integration and trajectory processing for RL pipelines, validating log probabilities, token masks, and loss inputs.
- Coordinate rollout and training workers, including checkpointing, restarting, and failure recovery, while tracking policy versions and sample staleness.
Requirements
- 5+ years of experience in ML systems, infrastructure, or distributed training.
- Experience modifying distributed ML frameworks (e.g., DeepSpeed, FSDP, FairScale, Horovod).
- Strong software engineering fundamentals, including Python, systems design, and testing.
- Proven multi-node experience (e.g., Slurm, Kubernetes, Ray) and debugging skills (e.g., NCCL/GLOO).
- Ability to implement algorithms across GPUs/nodes based on mathematical specifications.
- Experience working on an ML platform/infrastructure or distributed inference optimization team.
- Experience with large-scale machine learning workloads and strong ML fundamentals.
- Hands-on experience developing an LLM RL pipeline, covering rollout/inference and distributed training integration.
- Working knowledge of policy optimization methods (e.g., PPO, GRPO) sufficient for implementation and debugging.
- Familiarity with mixed-precision training (e.g., bf16, fp8) and accuracy validation.
- Experience with performance profiling, kernel fusion, or memory optimization.
- CUDA or Triton kernel experience.
- Experience with large-scale pre-training.
- Experience building and modifying custom training pipelines at scale.
- Deep familiarity with training infrastructure and performance tuning.
- Experience with asynchronous RL, multi-turn or tool-using rollout environments, or custom reward/verifier systems.