Lead Decision Intelligence; AI; NBA

The Lead Decision Intelligence Engineer (AI) owns the application of Decision Intelligence and agentic AI across the NBA platform. This role analyzes and formalizes the business decisions that drive member engagement, translating stakeholder objectives, constraints, policies, and available data into structured decision models that can be evaluated, optimized, and automated. Working closely with business, product, and engineering teams, you identify where decisions should remain rule-based, where predictive models should be applied, and where agentic systems can create measurable value.

You then design and build production-grade decision intelligence capabilities that help teams create, understand, optimize, and govern member actions. Using Lang Graph, Lang Chain, Azure OpenAI, Azure AI Foundry, Databricks, and Humana's AI Gateway, you build agentic workflows that reason through decision processes, generate recommendations, explain tradeoffs, assist with action authoring, and continuously improve decision outcomes. This is a hands‑on technical leadership role that combines decision science, AI engineering, and software architecture while leading a small team of engineers.

• Decision intelligence modeling — Analyze and formally model business decision processes, including objectives, constraints, policies, decision points, outcomes, dependencies, and feedback loops that govern member engagement.

• Decision decomposition — Break complex business processes into decision graphs, decision services, decision hierarchies, and optimization opportunities that can be measured, automated, and improved.

• Optimization strategy — Determine where rules, predictive models, reinforcement learning, optimization techniques, or agentic systems create the highest business value and operational impact.

• Agentic workflow delivery — Design and implement production agent workflows using Lang Graph and Lang Chain, including multi‑agent collaboration, tool usage, workflow memory, planning, reasoning, and human‑in‑the‑loop approval patterns.

• Action Library intelligence — Build AI-powered capabilities embedded directly into the Action Library that assist users in creating, refining, validating, governing, and optimizing member actions.

• LLM and agent engineering — Own integration with Azure OpenAI and other enterprise models through Humana's AI Gateway, including prompt engineering, structured outputs, retrieval patterns, tool calling, function execution, and workflow orchestration.

• Knowledge and retrieval systems — Design retrieval‑augmented architectures using vector search, semantic retrieval, knowledge grounding, and enterprise content sources to provide reliable decision context.

• Reinforcement learning integration — Partner with data science teams to operationalize reinforcement learning and decision optimization models within NBA workflows, ensuring recommendations can be deployed and governed at scale.

• Evaluation and experimentation — Build rigorous evaluation frameworks that measure recommendation quality, decision quality, agent effectiveness, user adoption, business outcomes, and operational performance.

• AI governance and safety — Implement guardrails, observability, traceability, policy controls, human review mechanisms, and auditability requirements appropriate for a healthcare environment.

• Team leadership — Lead and mentor AI engineers, establish engineering standards, conduct design reviews, and drive execution across the Decision Intelligence workstream.

• Cross‑functional partnership — Work closely with product, business, decision science, data science, and engineering teams to convert complex decision processes into production AI capabilities.