Senior Software Engineer, Sprig

• Build systems depth, including frontend: 6+ years owning a build system, with a deep understanding of build correctness and performance (hermeticity, sandboxing, caching, and invalidation). Hands-on experience owning production frontend build pipelines (webpack, Vite, esbuild, or Turbopack) and shared UI packages
• CI/CD as a platform: Experience owning incremental builds, test sharding, runner fleets, presubmit/postsubmit design, and merge queues, with an understanding that services and UI need different feedback loops
• TypeScript/Node fluency: Strong expertise in Node.js module resolution and packaging (CJS/ESM, exports)

Desirable

• Agents pay off when engineers hand them whole changes, which makes the build-and-test loop the safety gate for the entire org
• When CI is slow or flaky, teams lose time waiting and re-running to separate noise from real failures
• This role owns that loop: how quickly a change, from a human or an agent, becomes a signal people trust. That means incremental CI on a large TypeScript, React, and Go monorepo, fast and meaningful test automation, and short-lived environments spun up per branch instead of queued
• We're looking for a builder, not someone to inherit a ticket queue. You'll help set the direction for how development works in an agent-heavy world, then deliver it, as part of the Platform team that supports over a billion visitors a month and nearly a billion events a day!
• Modernize the JS/TS build: Lead the move to maintain Bazel tooling and replace accumulated workarounds with a cleaner, faster setup
• Make CI faster, cheaper, and observable: Track build and test performance, improve caching and parallelism, and give the team the data to make build decisions with confidence
• Make testing a first-class developer experience: Own how engineers and agents write, run, and trust tests, with local runs that match CI, automatic flake handling, and quality checks that act as real gates
• Build ephemeral, agent-ready environments: Replace shared, long-lived environments with fast, isolated, reproducible ones that people and agents can spin up on demand and use safely in parallel