Lead Genai Engineer (llm)

State Farm State Farm · Insurance · Bloomington, IL +3 · Technology and UX

State Farm is seeking a Lead GenAI Engineer to design and deploy LLM-powered applications, agentic workflows, and RAG pipelines using AWS Bedrock and AgentCore. The role involves building AI-integrated APIs, engineering data pipelines, championing responsible AI practices, and implementing DevSecOps for AI. The engineer will also enhance user experience through AI-powered interfaces and mentor team members.

What you'd actually do

  1. Design & Deploy AI/GenAI Solutions — Architect and implement LLM-powered applications, agentic workflows, & RAG pipelines using AWS Bedrock and AgentCore
  2. Build AI-Integrated APIs — Develop and maintain robust APIs (REST, FastAPI, etc.) serving as the backbone for AI applications, orchestration, and downstream consumers
  3. Engineer Data & Analytics Pipelines — Build pipelines that feed AI/ML systems with clean, governed, and well-structured data from enterprise sources
  4. Build on Amazon Bedrock AgentCore primitives — Runtime, Gateway (MCP), Memory, and Code Interpreter — to deliver streaming, stateful, tool-using agents in production.
  5. Champion Responsible AI — Apply prompt engineering best practices, implement guardrails, evaluate model outputs (Promptfoo, RAGAS), and ensure compliance with enterprise AI governance standards

Skills

Required

  • Python
  • TypeScript
  • AWS
  • Git
  • FastAPI
  • REST
  • prompt engineering
  • model evaluation
  • AI safety/guardrails
  • DevSecOps
  • CI/CD

Nice to have

  • AWS Bedrock
  • AgentCore
  • multi-agent orchestration
  • tool-use frameworks
  • GitHub Copilot
  • insurance
  • financial services
  • healthcare
  • Contact Center applications
  • AI governance

What the JD emphasized

  • production
  • enterprise-grade applications
  • regulated industries

Other signals

  • LLM-powered applications
  • agentic architectures
  • cloud-native AI
  • AWS Bedrock
  • AgentCore
  • Responsible AI
  • prompt engineering
  • guardrails
  • model evaluation
  • DevSecOps for AI
  • observability