Software Engineer - Baseten for Labs

Baseten is a company that powers mission-critical inference for dynamic AI companies, and they are seeking a full-stack, product-minded engineer to join their Labs team. This role involves building products that help model labs and AI researchers ship and scale their models, with a focus on both backend systems and developer-facing interfaces.

Responsibilities:

• Take meaningful ownership of projects: from API design and backend implementation to frontend surfaces, rollout, and operation
• Build backend services with high reliability and clear SLOs — auth, rate limiting, quotas, metering, and multi-tenant isolation
• Ship developer-facing product surfaces: dashboards, onboarding flows, and self-serve tooling that reduce time-to-value
• Collaborate closely with design, product, and GTM to define and ship what labs and developers actually need
• Drive performance and reliability improvements through profiling, tracing, and load testing

Requirements:

• 4+ years building and operating production software, including at least some full-stack experience (backend-primary is fine, but you're comfortable touching the frontend)
• Demonstrated ability to take initiative and contribute beyond the spec — you think about the 'why' behind what you build
• Strong backend fundamentals: API design, distributed systems, observability, and operational rigor
• Comfort working across the stack: backend services, data pipelines, and user-facing product surfaces
• Strong written communication — clear design docs, effective async collaboration
• Genuine curiosity about the AI/ML infrastructure space; you don't need ML expertise, but you want to understand the ecosystem
• Experience building developer-facing products: APIs, SDKs, CLIs, dashboards, or self-serve workflows
• Experience with API gateways, auth systems, billing/metering infrastructure, or multi-tenant platforms
• Frontend experience (React/TypeScript) or strong product UX instincts for developer tools
• Familiarity with model serving, LLM runtimes, or inference platforms
• Comfort with Kubernetes, distributed scheduling, or service mesh concepts