Senior AI Engineer, MapGPT
Core
Design, build, and evaluate multi-component AI systems (LLMs, agents) for Mapbox's navigation, search, and platform products, ensuring reliability, latency targets, and continuous improvement via evaluation frameworks.
Role type
Senior AI Engineer (LLM/Agent Systems)
Builds
Production AI systems, data pipelines, evaluation harnesses, and internal tooling for navigation and location intelligence.
Domain
Geospatial technology, AI/ML, LLMs, Agent Orchestration
Deliverable
production ML models
Required skills
LLM-backed feature shipping, data engineering (SQL, distributed processing), tool calling and agent orchestration, Python/TypeScript, evaluation of non-deterministic systems, distributed latency diagnosis
Preferred skills
Geospatial data (routing, geocoding, POI), MCP/transport experience, CI evals, automotive/infotainment, voice pipelines (ASR/TTS), offline/intermittent connectivity
Technologies
Python, TypeScript, SQL, LLMs, Agents, MCP, OpenStreetMap
Responsibilities
Define product behavior and build measurement/evaluation frameworks for non-deterministic systems; run continuous evaluation from end-user perspectives; own MVP delivery balancing technical perfection with shipping; build data pipelines (ingestion, conflation, entity resolution); track and evaluate external datasets/models; design feedback loops and failure instrumentation; design model boundaries and harnesses; optimize for latency and cost; build internal tooling (CLI, MCP); conduct code/design reviews and mentor team; participate in on-call rotation.