Responsibilities

We're looking for an exceptional Senior Software Engineer to help build the distributed backend that powers our AI agent platform and the developer-facing products around it — the event-driven services, real-time integrations, and APIs that let agents reliably do real work. You'll be building the systems that need to stay correct and available on our path to 1M users over the next year. You will:

• Build event-driven, horizontally-scalable backend services that process work reliably and correctly even when parts of the system fail.
• Build real-time surfaces and integrations — streaming connections, gateways, and CLIs — that hold up under load.
• Integrate LLMs and agentic harnesses into production systems.
• Drive features end-to-end — from ideation and architecture through deployment, observability, and iteration.
• Raise the engineering bar through clean, well-tested code, thoughtful reviews, and mentorship.

Requirements

Qualifications / Experience / Technical Skills

• 5+ years shipping and operating production backend systems at scale.
• Strong proficiency in Python, and the judgment to write clean, well-tested, maintainable code under it.
• Deep distributed-systems fundamentals: you know how to keep systems correct and available under partial failure, backpressure, and load — and you design for it by default.
• Hands-on experience with event-driven, streaming, or real-time systems (queues, pub/sub, long-lived connections).
• Comfortable operating services in production on Kubernetes in the cloud — and instrumenting them with metrics, tracing, and dashboards so you can debug what you ship.
• Solid API and software-architecture instincts; you make and defend sensible trade-offs in ambiguity.
• LLM/agentic and developer-facing product experience (MCP, RAG, APIs, SDKs, CLIs) a strong plus.

Our stack: Python, FastAPI, Postgres, Kubernetes, AWS. You don't need prior experience with all of it — we care that you can pick it up fast.

Soft Skills / Personal Characteristics

• High agency and ownership: You drive work end-to-end, from ambiguity to shipped and iterated, without waiting to be told.
• A reliability mindset: You design for failure, not just the happy path.
• Product-minded: You care about the users using what you build, not just the system behind it.
• AI-native: You use modern AI tooling (e.g., Claude Code) in your daily workflow and stay at the cutting edge of what it can do.
• Excellent written communication: We work async and write things down.