[Remote] Staff AI Engineer, Agents and Evals

Note: The job is a remote job and is open to candidates in USA. MimoLabs builds Mimo, an agentic system that helps people design and produce physical things. The Staff AI Engineer will own how the agent thinks and behaves, including its context and tools, and build the evaluations that determine what changes ship.

Responsibilities

• Own the context layer: what Mimo knows, when, and at what cost, with system context, retrieval and memory managed as one budget, for example the moment a user's own material library enters the conversation
• Define the tool contracts: the definitions, results and failure modes of the tools Mimo calls, for example what a manufacturability check returns when the part cannot be made
• Build the evaluation practice: offline suites, graded live passes and regression gates for every behaviour change, for example a benchmark of real requests scored for the part they produced
• Shape how Mimo reasons, asks, explains and recovers, and the rules users set for their own Mimo
• Decide which model runs where, and the gate a new one must pass before it ships

Skills

• Agent systems in production : tool use, multi-step loops, streaming and failure recovery, with results you can show
• Evaluation design : graded suites, regression gates and human review that you built before trusting a change
• Context engineering : prompts, retrieval and memory managed as one budget, with the costs measured
• Tool design : contracts a model can use reliably, with schemas, results and errors
• Python
• A large production codebase with a strict test baseline or a published benchmark you built
• Clear writing : an evaluation report a founder can act on
• Fluency with agent systems matters most for this role; you already build this way
• US work authorization is required, and we are not able to sponsor visas for this role
• Parametric or code-first CAD authoring; 3D geometry fluency
• Multi-model routing and cost attribution; fine-tuning for tool use
• Retrieval systems in production
• Published benchmarks or tool-use evaluations
• Voice and UX writing for an agent product

Benefits

• Equity, four-year vesting with a one-year cliff
• Health, dental and vision
• Remote across the US, with occasional travel to Newlab in Detroit

Company Overview