An Artificial Intelligence Engineer develops, tests, and deploys artificial intelligence models. May work closely with data software engineers and data professionals to train and implement AI models into existing systems or develop new applications.
Lead AI Engineer Carzato Detroit, MI Job Details Full-time | Contract $140,000 - $160,000 a year 10 hours ago Qualifications AI models Data model design Systems integration Software engineering Data modeling projects System design Public Cloud System design for system development Production systems AWS Model deployment Distributed systems APIs Developing and maintaining backend systems Distributed computing AI-driven automation Machine learning libraries Machine learning frameworks Design (software development lifecycle) Full Job Description About Carzato Carzato builds the digital retailing software that powers how people actually buy cars online. For eight years, our platform has run on dealer and OEM websites with configuration, financing, trade-in valuation, and new vehicle reservations. Today it operates across more than 1,000 dealerships. Our next release rebuilds that experience around
AI:
natural-language vehicle research and conversational search, grounded in live inventory, pricing, and incentive data. We've done the research, modeled the economics, and set the architecture. It launches early next year. That's the distribution problem most AI products never solve, and we already have 1,000+ sites ready to run it on day one. What we need now is the engineer who builds it and the platform underneath it. The Short Version Carzato has been building automotive digital retailing software for eight years. Our next release is AI-native, conversational vehicle research and search grounded in live inventory and pricing data. It's in active development now, launching early next year, and it ships across our entire dealer network at once. We're hiring a Lead AI Engineer to build it and own the platform beneath it. What You'll Actually Do Design and build RAG pipelines against inventory and pricing data that changes daily , such as retrieval, chunking, grounding, re-ranking Own the AI platform architecture: model routing, provider integration, caching, fallback behavior Build the backend services, APIs, and data infrastructure AI features depend on Define evaluation and monitoring so we catch quality drift before dealers and shoppers feel it Manage cost and latency as first-class constraints , such as our volume, per-session economics decide what's viable Write specs and coordinate delivery with our offshore development team Work directly with product, design, and leadership to move from validated direction to shipped product
What We Need From You Non-negotiable:
3-5 years shipping production software you were accountable for Strong proficiency in Python and/or TypeScript Real backend and system design depth: APIs, data modeling, distributed systems Experience with a major cloud platform (AWS, GCP, or Azure) Production experience integrating LLMs or ML models into real applications Working knowledge of
RAG:
embeddings, vector search, retrieval quality, grounding Daily, substantive use of AI coding agents (Claude Code, Cursor, Copilot) in real work; you can describe specifically where they help and where they mislead Comfort operating in a fast-moving environment, and the judgment to know when to ask versus decide Strong plus: LLM evaluation frameworks and systematic output quality measurement Agentic workflows and multi-step task automation Vector databases (Pinecone, Weaviate, Chroma, pgvector) Cost and latency optimization at production scale Automotive, e-commerce, or marketplace experience ML frameworks (PyTorch, TensorFlow) CS, Software Engineering, or related degree or an equivalent body of shipped work
Why This Seat:
You arrive before launch. Most engineers join AI products already built by someone else. This one is yours from v1. Distribution is already solved. Over 1,000 dealer sites. Your work ships to the full network ; no growth hacking, no waiting for users. Technical ownership of a domain, not a ticket queue. You set the AI architecture. No platform team to negotiate with, no committee between you and the decision. Eight years of stability, greenfield engineering. Established company, real revenue, proven platform; and the AI stack is a blank page. Direct line to leadership. You report to the VPs and Leadership. Short path from idea to decision. Remote, with real autonomy. We measure output, not hours. How We Work Small, senior team. High autonomy, short feedback loops. We ship a working v1, put it in front of users, and iterate. We care about what's in production, not what's in the deck. We care more about engineering depth and demonstrated AI fluency than a specific credential list. If you're strong on fundamentals and you've been building seriously with LLMs, we want to talk.
