AI Engineer

Location: San Francisco, CA Type: Full-time Tags: Engineering

We're a Sequoia-backed stealth startup building the operational memory layer for autonomous engineering. Cursor and Claude Code take you from idea to code. We take you the rest of the way — through infrastructure, across teams, across providers — without agents needing to hand back code/operations to a human. And we do this at both design and run time. We're live with enterprise customers, freshly funded, and moving fast.

We're looking for AI Engineers to build the intelligence layer on top of our infrastructure graph — turning structured change data into autonomous reasoning, proactive risk detection, and agent-native tooling.

• Design and deploy AI-driven agentic systems that reason over live infrastructure graphs to detect blast radius, predict risk, and surface causal change chains.
• Partner with product and engineering to translate emerging LLM research into customer-facing capabilities — particularly around AI coding agents that need grounded, accurate infrastructure context.
• Rapidly prototype and iterate on AI features based on direct customer feedback from production deployments at enterprise accounts.
• Own reliability and scalability of AI systems operating at enterprise scale — this is not a research role, your work ships.
• Drive integration of graph-native retrieval, MCP tooling, and structured context delivery into agent workflows.

• 2 years of hands-on AI/ML engineering experience, with a CS, Engineering, or related degree.
• Strong track record applying LLMs to complex, unstructured problems — ideally in infrastructure, DevOps, or developer tooling domains.
• Self-directed and comfortable operating in ambiguity — we're early, the problems are novel, and there's no playbook.
• Clear communicator who can move between deep technical implementation and customer-facing discussion without losing altitude.
• Familiarity with graph data models, knowledge graphs, or infrastructure observability is a meaningful plus.

We're in-office 5 days a week in San Francisco.

We're in stealth but moving. If you want to work on a problem that sits at the intersection of infrastructure intelligence and the AI agent stack — reach out.