About the role
- Design and build LLM serving infrastructure on Kubernetes: deployment, GPU scheduling, scaling, and model lifecycle management
- Package the platform for enterprise environments: Helm-based installs, upgrades, and restricted/offline networks
- Integrate the serving layer with the platform’s API gateway, identity, and metering services
- Build the observability for operating GPU inference in production (serving metrics, GPU telemetry)
- Contribute across a multi-service codebase and help set engineering direction through design docs and reviews
- Deep hands-on Kubernetes experience: building and operating production workloads and Helm charts, not just consuming managed clusters
- Strong Go programming skills; solid CI/CD and infrastructure-as-code skills
- Experience with GPU workloads or LLM inference, or strong adjacent systems experience and a track record of learning fast
- 5+ years of software engineering experience in infrastructure, platform, or distributed systems
- Comfortable with high autonomy on a small, remote-first, written-culture team
- Fluency with AI-assisted development tools (Claude Code, OpenAI Codex) as part of your daily engineering workflow
- Inference performance work (quantization, batching, caching) or distributed serving frameworks
- Enterprise deployment experience: air-gapped installs, SSO/OIDC, supply-chain security
- UI development experience (e.g. React/TypeScript), useful as the product’s management surfaces grow
- Open-source contributions in the Kubernetes or ML-infrastructure ecosystems