ML Infra Engineer
About the role
You will help scale and optimize our training systems and core model code. You’ll own critical infrastructure for large-scale training, from managing GPU/TPU compute and job orchestration to building reusable and efficient JAX training pipelines. You’ll work closely with researchers and model engineers to translate ideas into experiments—and those experiments into production training runs.
Responsibilities
- Own training/inference infrastructure: Design, implement, and maintain systems for large-scale model training, including scheduling, job management, checkpointing, and metrics/logging.
- Scale distributed training: Work with researchers to scale JAX-based training across TPU and GPU clusters with minimal friction.
- Optimize performance: Profile and improve memory usage, device utilization, throughput, and distributed synchronization.
- Enable rapid iteration: Build abstractions for launching, monitoring, debugging, and reproducing experiments.
- Manage compute resources: Ensure efficient allocation and utilization of cloud-based GPU/TPU compute while controlling cost.
- Partner with researchers: Translate research needs into infra capabilities and guide best practices for training at scale.
- Contribute to core training code: Evolve JAX model and training code to support new architectures, modalities, and evaluation metrics.
Requirements
Strong software engineering fundamentals and experience building ML training infrastructure or internal platforms. Hands-on large-scale training experience in JAX (preferred), PyTorch. Familiarity with distributed training, multi-host setups, data loaders, and evaluation pipelines. Experience managing training workloads on cloud platforms (e.g., SLURM, Kubernetes, GCP TPU/GKE, AWS). Ability to debug and optimize performance bottlenecks across the training stack. Strong cross-functional communication and ownership mindset.
Qualifications
Deep ML systems background (e.g., training compilers, runtime optimization, custom kernels). Experience operating close to hardware (GPU/TPU performance tuning). Background in robotics, multimodal models, or large-scale foundation models. Experience designing abstractions that balance researcher flexibility with system reliability.