Senior AI Engineer

About the Role:

• You'll own agents end to end within a problem area, from first prototype to measured production performance.
• You'll join a new engineering group building AI-native products for auto casualty business, specifically the injury side of auto claims.
• The group is small, senior, and growing fast, working on medical bills, claims decision support, and an Agentic Rules Platform.

Key Responsibilities:

• Design and build LLM-powered agents and workflows: orchestration, tool use, retrieval, structured extraction, and human-in-the-loop review.
• Build evaluations with data labelers and subject matter experts, run them on every change.
• Make agents production-grade: guardrails, observability, cost and latency, and careful handling of sensitive health data.

What success looks like:

• In your first month, you've shipped a measured improvement to an existing agent.
• By six months, an agent you built is in production with customers, backed by evaluation results.
• You'll own the orchestration, retrieval, evaluations, and guardrails that make agents safe to rely on.

Requirements:

• Typically 5+ years of professional software or ML engineering experience, including meaningful recent work with LLMs.
• Strong production engineering skills in Python, including API services, and comfort working in TypeScript.
• At least a year of shipping LLM-powered products to real users, including retrieval (RAG, embeddings, vector search) and prompt engineering.