Remote Senior AI Engineer, MapGPT

What you'll do: This role is scoped by skill rather than by product. The problems below span multiple departments and show up across our Search and Places data work, our Location and Navigation Intelligence work, and the Platform work that makes Mapbox usable by agents. You'll be hired into one specific team or product area, but you'll work across teams where your expertise meets the highest-priority AI problems. You should expect to move between teams and tech stacks as the work demands. At this level you own the technical design and delivery of a multi-component AI system, and you are accountable for the quality of what ships in your area.

In this role, you will: Define how the products you own should work, build a measurement framework for it, and build evaluation systems for non-deterministic behavior. Formulate hypotheses around the products we build and seek the signal needed to validate them. Define what a correct result is for a given input and state, determine whether to assemble datasets from real usage or hand-written cases, and gate changes on regression results. Run continuous evaluation of the products we build, whether APIs, SDKs, data representations, or reference applications, from the position of the end user, whether developer, agent, or consumer. Assess the gaps such as misuse of parameters or integration anti-patterns, recommend the fixes, and make sure they land. Own the MVP against an agreed north star technical design, and balance technical perfection against shipping useful increments. Build data pipelines and the tooling around them: ingestion, conflation, entity resolution, quality checks, and the batch and streaming jobs that keep a large dataset current. Track and pull external datasets, models, and benchmarks from published research and open-source releases. Evaluate what fits the problem and constraints, and decide when to adopt what exists versus build your own. Design feedback loops so that using a product generates data that improves it. Instrument systems so failures arrive with enough context to reproduce, then turn the recurring ones into evaluation cases. Design the boundary between a model and the tools it calls. Build or improve the model harness, decide what the model handles, what it delegates, and how to keep it working from the state it actually fetched. Work to a latency and cost target per request: streaming, partial results, caching, model routing, prompt structure. Build the internal harnesses and tools (CLI, MCP, and others) your team needs to iterate quickly, and share the parts that generalize with other teams. Raise the bar on your team through code and design review, and bring other engineers up on eval practice. Some of the technical questions in this area are still open. You will help answer them. Participate in an on-call rotation to ensure our systems remain available to customers 24/7. Team members alternate as the on-call primary responder, which may require immediate response outside normal working hours, including weekends. What We Believe are Important Traits for This Role