Y Combinator Software Engineer
Summary
Build AI agent loops for mobile app testing—agents that observe screens, execute actions, and reflect on results—along with distributed workflow orchestration, device infrastructure, a CLI, and real-time streaming UX.
What you'd work on
- The AI agent loop. Our agents observe device screens, ground actions to UI elements, execute steps, and reflect on whether they worked. It's a capture-ground-execute-reflect cycle and there are a lot of hard problems in making it reliable
- Distributed workflow orchestration for test execution across pools of devices
- Device infrastructure for Android and iOS. Emulators, real devices, streaming, automation
- Our CLI that ties local dev to cloud verification. Build uploads, live device sessions, CI integration
- OpenTelemetry integration, trace correlation, and the tooling we use to debug executions
- Real-time execution UX. WebSocket streaming, live device views, graph-based test visualization
- Building out product surface (CLI, web app, CI integrations)
What we're looking for
- High customer empathy. You think about the user's problem when you're designing a technical solution, not just the technical elegance
- You enjoy novel problems with real technical and product ambiguity
- You've built and scaled systems that run in production under real load
- Practical experience with LLMs, AI infra, or agent-based systems
- Track record of high ownership and autonomous execution
- Strong communication
Experience that would stand out
- Mobile build tooling experience (Swift, Kotlin, Flutter)
- Mobile testing infrastructure, device automation, or mobile platform tooling
- OpenTelemetry or distributed tracing
- Workflow orchestration systems
- CLI or developer tooling
Probably not a fit if
- You want well-scoped tickets
- You're uncomfortable making architectural decisions without full information
- You prefer deep specialization in one layer over working across a full stack
Why join?
- Hard technical problems (AI agents, distributed systems, mobile device infrastructure) not CRUD apps
- Shape technical and product direction from the ground floor
- Flat hierarchy, high ownership, full autonomy
- Small, talent-dense founding team in San Francisco
- Competitive compensation and equity