Senior AI Engineer

What You'll Do

• Design, build, and ship production LLM powered features end-to-end. 
• Build the AI application layer, including prompt orchestration, tool calling, retrieval (RAG), embeddings, agent workflows, and structured outputs.
• Develop evaluation infrastructure, including regression suites, LLM-as-a-judge evaluations, synthetic datasets, human review workflows, and quality metrics that catch regressions before users do.
• Design observability systems that measure accuracy, latency, cost, hallucination rates, and model behavior in production.
• Build safety and reliability guardrails that ensure AI systems meet quality, compliance, and operational requirements.
• Evaluate and integrate third-party tools, including orchestration frameworks, evaluation platforms, vector databases, and model providers; and make pragmatic build-versus-buy decisions.
• Raise engineering standards across the team through technical leadership, code reviews, mentorship, and thoughtful system design.
• Sit with EHR operators, Patient Advocates, and clinical teams to understand the workflows you're automating.
• Prototype quickly, ship early, measure outcomes, and iterate based on real-world usage.

What You’ll Bring to the Team

• 5+ years building production software, with at least 1–2 years building and deploying LLM-powered applications in production (or equivalent depth through substantial side projects or open-source work).
• Strong Python engineer with experience building backend services and integrating modern AI APIs and frameworks.
• Comfortable using AI-assisted development tools throughout the software development lifecycle.
• You've shipped AI features that served real users, and you can clearly articulate how you measured quality, reliability, latency, cost, and business impact.
• You have strong product instincts and thrive in ambiguous environments, translating operational problems into technical solutions without relying on detailed specifications.
• Bonus: Experience with orchestration frameworks (LangChain, LangGraph, AgentCore), RAG architectures, vector databases, prompt/version management, and evaluation or observability platforms (e.g., LangSmith, Braintrust, Arize Phoenix).