2 months ago
Seattle, WA, USA +2 moreSenior / Staff+
H1B Sponsor
Base Salary
$320k - $485k/yr
Responsibilities
- Own unattended deployment orchestration for validated inference builds across GPU, TPU, and Trainium fleets.
- Improve capacity-aware scheduling to maximize deployment throughput under constrained accelerator budgets and variable fleet sizes.
- Extend deployment observability with dashboards and tooling that show production code, commit progress, and validation results.
- Reduce merge-to-production cycle time through pipeline architectures that minimize serial dependencies and maximize parallelism.
- Optimize fleet rollout strategies across thousands of accelerator chips while minimizing disruption to serving capacity.
- Evolve self-service model onboarding so new models can enter the continuous deployment pipeline without Launch Engineering involvement.
- Partner with validation, autoscaling, model-routing, and infrastructure teams to integrate deployment automation.
Requirements
- Strong software engineering skills designing systems with complex state machines and multi-stage pipelines.
- Proficiency with Kubernetes-based deployments, rolling update mechanics, and container orchestration.
- Experience building deployment, release, or delivery infrastructure shaped by resource constraints.
- Track record of automation that measurably improves deployment velocity and reliability.
- Ability to work across backend services, databases, CLI tools, and web UIs.
- Strong communication skills and ability to collaborate with on-call engineers, model teams, and infrastructure partners.
- Preferred: 5+ years building deployment, release, or delivery infrastructure at scale.
- Preferred: Production experience with Python and/or Rust.
- Preferred: Experience deploying ML inference or training infrastructure across GPU, TPU, and Trainium.
- Preferred: Capacity planning or resource-constrained scheduling experience, such as bin-packing, fleet management, or hardware-affinity job scheduling.
- Preferred: Progressive delivery experience including canary or soak testing, blue-green deployments, traffic shifting, and automated rollback.
- Preferred: Experience with large-scale release engineering challenges such as mobile release trains, monorepo deployments, or multi-datacenter rollouts.
- Bachelor’s degree or equivalent combination of education, training, and/or experience in a relevant field.
Benefits
- Hybrid policy requiring staff to be in an office at least 25% of the time
- Visa sponsorship available with immigration-lawyer support
- Competitive compensation and benefits
- Optional equity donation matching
- Generous vacation and parental leave
- Flexible working hours
- Collaborative office space
Tech Stack
About Anthropic
We're an AI research company that builds reliable, interpretable, and steerable AI systems. Our first product is Claude, an AI assistant for tasks at any scale. Our research interests span multiple areas including natural language, human feedback, scaling laws, reinforcement learning, code generation, and interpretability.