AI Engineer

Why Mobility Global? At Mobility Global, our mission is to provide the trusted information that helps billions of people build, sell, and own vehicles with confidence. As we shape the next generation of automotive intelligence through innovation, AI, and industry-leading data, we're creating new opportunities for our customers, our business, and our people. Join a team singularly focused on transforming the automotive industry. S&P Global Mobility is Mobility Global — the company behind CARFAX, automotiveMastermind, Market Scan, and Polk Automotive Solutions. Our strength comes from our people, and here, the progress you make helps the entire automotive ecosystem move faster and with greater confidence. The Team: You will join the AI Engineering team at Mobility Global. We build the conversational analytics products, shared AI services, and agent platforms that sit on top of the industry’s deepest automotive dataset, and we build the internal tooling that drives AI adoption across the PDLC/SDLC, from requirements and design through code, testing, and release. We are a small, senior, product-minded team that owns its work end to end, from the React front end a customer touches through the FastAPI services and data layer behind it to the evaluation harnesses that tell us whether the answer was right. We work AI-natively by default: coding agents, agentic workflows, and LLM-assisted review are part of how the team ships every day, not a side experiment, and we care as much about engineering judgement as we do about raw output. This role is a deliberate early-career seat on that team, with senior engineers close by and real production ownership from the first quarter. Responsibilities and Impact: Build LLM-powered features in Python — agentic workflows in LangGraph, retrieval and RAG pipelines, text-to-SQL, and tool/MCP integrations that reach real users rather than staying in a notebook. Ship the services behind them — write clean, typed, tested Python and FastAPI endpoints, including streaming responses, async I/O, and sensible error handling. Own evaluation and quality for what you build — assemble golden datasets, write eval and guardrail test suites, instrument traces, and use the results to decide whether a change actually helped. Do prompt and context engineering as an engineering discipline — iterate on prompts, tool schemas, and retrieval strategy, and measure the difference instead of guessing at it. Debug agent behaviour end to end — read traces, isolate failures across retrieval, tool calls, parsing, and entitlements, and turn each one into a regression test. Work daily with AI coding agents and with the people around you — use agents to move faster while reviewing their output carefully, and pair with senior engineers, data scientists, and product partners to learn the domain quickly. What We’re Looking For: Basic Required Qualifications: 1 to 5 years of professional software engineering experience , with a…