The roleYou'll own the backend behind our AI features. That's the APIs, the data pipelines, and the orchestration sitting between our product and the language models doing the work. In practice it means running retrieval over large sets of documents, getting dependable results out of LLMs and their tool calls, streaming responses to the UI, and processing long jobs in the background without things falling over.
It isn't CRUD work. The hard parts here are keeping AI systems accurate and fast with confidential data, at production scale. You'll make architecture calls early and live with them.
RequirementsWhat You'll Do- Designing and building the APIs and services our products run on.
- Working on the AI plumbing: retrieval, orchestration, tool-calling, response streaming, and the evals that tell us whether the output is any good.
- Keeping async and distributed work reliable: document ingestion, embeddings, and long-running jobs on Redis.
- Building integrations with the systems clients already use, like email and document stores, and making them hold up under load.
- Taking data security and isolation seriously, since clients hand us confidential material.
Who You Are- At least seven years building backend systems that run in production. Depth matters more to us than the number.
- You think about how things fail, not just how they behave when everything's fine.
- You've built real LLM-powered systems, or you're itching to: RAG, tool-calling, agents. You know where they tend to break.
- You use AI coding tools (Claude Code, Cursor, Copilot) and use them well. Faster output, but you understand and stand behind every line; you wouldn't approve sloppy AI-written code in review, so you don't ship it either.
- You're fine in a startup that changes direction. Decisions get made, then revisited.
Experience we expect with the stack:Nice to have: RAG or vector search experience; LangChain or LlamaIndex; streaming over SSE, WebSocket or SignalR; Microsoft Graph and email integration work.
Benefits 
