AI Engineer — Applied AI

About Us Savant is transforming how healthcare and life sciences organizations unlock the value trapped in unstructured medical data. Our platform combines cutting-edge large language models (LLMs) with domain-specific quality controls to convert free-text clinical records into structured, analysis-ready data — efficiently, accurately, and at scale. We work with leading institutions across healthcare, life sciences, and research, supporting faster studies, sharper insights, and better care. Backed by Roivant (NASDAQ: ROIV), Savant is built for organizations that see structured data not just as an output, but as a foundation for innovation. The Opportunity We’re hiring an AI Engineer to improve the quality and efficiency of Savant’s clinical data platform. This is a generalist applied-AI role for someone who can move fluidly between software engineering, data science, experimentation, and production operations. You’ll study how changes throughout our pipeline affect system performance, then design and ship improvements. That may mean changing how we prompt or structure interactions with LLMs; how we retrieve, rank, and assemble supporting information; how workflows are decomposed and controlled; or how outputs are validated and repaired. If you’re excited by the challenge of making AI systems measurably better, enjoy following evidence across the entire stack, and want your experiments to become real product capabilities, we’d love to hear from you. Savant is based in NYC, and this role is remote-eligible. Preference will be given to candidates able to come to work in NYC. Some Things You Might Work On Establish software programming paradigms for our platform’s interactions with LLMs in both token-parsimonious and token-hungry scenarios Build a test bench for understanding the $/performance frontier of open- and closed-weight language models as relates to our platform’s (often very different) core workloads Partner with clinical experts to encode domain knowledge and define meaningful quality standards Build evaluation suites that measure accuracy, completeness, consistency, latency, and cost Design controlled experiments that isolate the effects of changes throughout the platform Improve model interactions through prompting, structured outputs, tool use, context management, and workflow composition Build datasets and internal tools for evaluation, regression testing, and rapid iteration What We’re Looking For 5+ years of experience in applied machine learning, AI engineering, data science, or adjacent software engineering roles Excellent Python and SQL skills Hands-on experience building production applications with LLMs Experience designing experiments and interpreting noisy, multidimensional results Familiarity with LLM evaluation, retrieval-augmented generation, agentic workflows, or MLOps Strong software-engineering fundamentals and the ability to ship reliable production code Intellectual…