About the role
You will build and operate ML systems for model training, evaluation, data ingestion, experiment orchestration, and serving. You will improve distributed training, develop large-scale data pipelines, operate workflow and inference systems, and maintain deployment, observability, security, and CI/CD capabilities.
Responsibilities
- Own ML systems for training, evaluation, and serving foundation models
- Improve distributed training performance, stability, reproducibility, and checkpointing
- Build Python data ingestion, preprocessing, validation, versioning, and publishing pipelines
- Operate workflow orchestration, GPU scheduling, experiment tracking, and inference systems
- Ship workloads using Docker, Kubernetes, Terraform, and CI/CD pipelines
- Implement monitoring, logging, alerting, SLOs, and incident response
- Manage secrets, IAM, and network boundaries
- Partner with researchers, engineers, and the platform team to improve ML systems
Requirements
- 3+ years writing production-quality Python in a large multi-author codebase
- Experience with PyTorch distributed training stacks and debugging multi-GPU jobs
- Experience shipping end-to-end data pipelines at scale
- Experience with GPU compute and performance debugging
- Working knowledge of AWS, GCP, or Azure
- Proficiency with Docker, Kubernetes, and Terraform
- Knowledge of SQL, relational, analytical, embedded, and object storage systems
- Familiarity with workflow orchestration, experiment tracking, observability, and CI/CD tools
Requirements
- 3+ years writing production-quality Python in a large multi-author codebase
- Experience with PyTorch distributed training stacks and debugging multi-GPU jobs
- Experience shipping end-to-end data pipelines at scale
- Experience with GPU compute and performance debugging
- Working knowledge of AWS, GCP, or Azure
- Proficiency with Docker, Kubernetes, and Terraform