Systems ML Engineer: Cloud & Edge Performance

S27a; Cambridge (MA); On-site; USD 170,000 - 240,000; Full time

Benefits offered by this job

Competitive compensation

Full benefits

Job summary

Transfyr in Cambridge, MA is seeking a Systems ML Engineer to optimize and deploy large-scale ML models across training and production. You will own performance, profile bottlenecks, and implement kernel-level improvements across cloud and edge environments.

You will collaborate with research and perception teams to push training efficiency, manage deployments with monitoring and rollback, and ensure robustness of multimodal data pipelines in real lab settings.

Qualifications

• Proven ML systems engineering in production environments.
• Experience with distributed training and model serving.
• Cloud & edge infrastructure proficiency (AWS, Kubernetes).
• Understanding of ML stack from hardware to deployment.
• Strong cross-stack debugging: GPU, data loading, memory, and communication.
• Experience with CUDA/kernel tuning for performance.

Responsibilities

• Profile and optimize performance using Nsight, PyTorch Profiler to identify bottlenecks and implement kernel-level improvements.
• Improve efficiency of distributed training pipelines with PyTorch Distributed.
• Develop and maintain high-performance GPU kernels in CUDA/Triton for critical workloads.
• Design and optimize data loading and inference pipelines for multimodal lab data.
• Manage deployment across cloud and edge in active lab environments with attention to security and cost.
• Debug and resolve production bottlenecks, resource issues, and failures across the stack.
• Collaborate with research and perception teams to move data through pipelines reliably.
• Implement monitoring, versioning, and rollback in deployments to protect updates.

Skills

ML systems engineering

Distributed training

Cloud & edge infra

Full ML stack

Cross-stack debugging

Algorithm optimization

Tools

CUDA

Triton

PyTorch

Kubernetes

Terraform

NCCL

Job description

Transfyr in Cambridge, MA is seeking a Systems ML Engineer to optimize and deploy large-scale ML models across training and production. You will own performance, profile bottlenecks, and implement kernel-level improvements across cloud and edge environments.

You will collaborate with research and perception teams to push training efficiency, manage deployments with monitoring and rollback, and ensure robustness of multimodal data pipelines in real lab settings.