AI Engineer

Role overview

A small consultancy is hiring a senior AI engineer to take one scoped client problem from a blank repository to a production system in 30 to 60 days, then stay engaged when things break. The work closes the gap between an impressive demo and a system an operations team can trust on a bad day, with evals, monitoring, and escalation paths designed in from the start.

Responsibilities

- Drive a scoped engagement from discovery through production, owning architecture calls end to end. - Write the eval set before any build, with real inputs running in CI so prompt or model changes are measured rather than guessed. - Build a reliability layer covering structured outputs, retries, fallbacks, confidence thresholds, and human escalation where the risk warrants it. - Implement governance controls including access boundaries, tenant isolation, audit trails, and review checkpoints that matter most in regulated work. - Develop MCP servers and tool schemas that external agents can integrate against. - Hand off clean, well-documented systems the client's engineers can extend without ongoing help.

Requirements

- Three or more years shipping real software, with backend depth in TypeScript/Node.js or Python. - Production LLM experience across prompting, context engineering, agent architectures, retrieval, and evals, including systems that kept working after launch. - Ability to take an empty repository to a deployed system without waiting for a written spec. - Comfort engaging directly with clients and holding a technical position without hiding behind jargon. - Strong written and spoken English.

Nice to have

- Background in regulated environments such as health or finance. - Hands-on experience with MCP, n8n, retrieval at scale, or Next.js.

Benefits and work setup

- Fully remote with flexible hours. - Direct access to leadership and clients with no ticket-queue layers. - Generous hardware budget. - Stated room to grow as the company grows.