2 days ago
Vancouver, CanadaMid Level
Responsibilities
- Scale AI products on cloud GPU infrastructure while managing performance, reliability, and cost.
- Build end-to-end model deployment pipelines for packaging, versioning, staged releases, and rollback.
- Set up monitoring and alerting for uptime, latency, and model performance, and respond to production incidents.
- Provide reliable training infrastructure, reproducible pipelines, and versioned data and models for AI engineers.
- Coordinate releases and issue triage with CAD design and QC teams.
- Create runbooks and handoff documentation for production systems.
Requirements
- At least 3 years of experience as a software engineer running production cloud services, ideally on Azure, with responsibility for uptime and operating costs.
- Hands-on experience deploying machine learning models to production using a serving framework such as Triton, TorchServe, Ray Serve, or SageMaker, ideally on GPUs.
- Experience with Docker, Kubernetes, infrastructure as code such as Terraform, and versioned releases with rollback.
- Experience with monitoring and alerting tools such as Prometheus, Grafana, or Datadog, including production incident response.
- Strong Python skills and the ability to read PyTorch code and understand computer vision fundamentals.
- Clear technical communication, documentation, and collaboration with non-engineering teams.
- Preferred qualifications include experience with 3D data or medical imaging, ONNX, TensorRT, batching, GPU autoscaling, MLflow, DVC, model registries, event-driven pipelines, regulated or healthcare environments, and distributed time-zone collaboration.
Benefits
- Onsite work five days per week at the Vancouver office at 564 Beatty St.
- Shared on-call coverage for Vancouver production support, including some after-hours responsibility.
- Annual base salary of $100,000–$120,000 CAD.
- Individual contributor role with no direct reports initially.
