| Location | Minneapolis, MN |
Role Overview Hands-on engineering role (prototype to production) combining strong Python and software engineering fundamentals with AI/ML, GenAI, RAG, and Agentic AI engineering.
Primary Responsibilities
Design, develop, test, and deploy Python-based AI/ML and Agentic AI solutions for DSNP business/analytics workflows.
Build intelligent agents using LLMs, RAG, tools/APIs, enterprise data, orchestration, and human-in-the-loop patterns.
Integrate AI capabilities with existing DSNP applications, APIs, analytics platforms, and enterprise data stores.
Develop production-grade AI services with robust testing, evaluation, monitoring, security, and observability.
Collaborate cross-functionally (Product, Data, UX, Architecture, Engineering) to translate business needs into scalable solutions.
Required Qualifications
Experience: 6+ years in software, AI/ML, or data engineering, including 4+ years of hands-on Python development.
Core Python: Advanced skills in APIs, data processing, unit/integration testing, debugging, and production-quality software development.
GenAI & Agentic AI: Hands-on development with LLMs, RAG, embeddings, vector search, prompt engineering, structured outputs, tool/function calling, multi-step workflows, memory/context management, and orchestration.
Integration & Systems: Integrating AI models with enterprise APIs, relational/vector databases, and cloud platforms.
Engineering Best Practices: Git, CI/CD, evaluation frameworks, monitoring, observability, security, and responsible AI.
Preferred Qualifications
Cloud & Frameworks: Azure AI/ML, Azure OpenAI, Semantic Kernel, LangGraph, AutoGen; FastAPI, SQL, containerized deployments.
Domain: Healthcare payer, Medicare Advantage, insurance, or regulated enterprise environments (PHI/PII handling).
AI Governance: Guardrails, hallucination mitigation, model evaluation, and compliance.
Initiatives & Expectations
Projects: DSNP Food Benefit Platform / Medicare Advantage Plan Optimization (decision support, RAG, agentic workflows).
Tooling: 90%+ weekly utilization of enterprise AI tools (GitHub Copilot, M365 Copilot) across development, testing, and documentation.