Principal Product Manager, AI

Humana•Washington, DC

1d•$201,600 - $277,200•Hybrid

About The Position

As Humana continues its transformation into an AI-first enterprise, we are looking for a hands-on, technically deep Principal Product Manager to define strategy and drive delivery across a portfolio of AI products. You will set the standard for how we build AI at Humana, and personally go deep, from understanding a business challenge, to prototyping an AI solution, to proving its value, to delivering it through Humana's enterprise AI governance process. You will lead through influence, making engineers, data scientists, designers, product managers, and business leaders more effective.

Requirements

10+ years of product or related experience, with AI/ML or GenAI products shipped to production. Candidates with fewer years but a strong engineering background and real fluency in AI will be considered.
A track record of setting strategy and driving delivery across multiple products or a major domain, and of leading initiatives with org-wide, executive-level visibility.
You use AI tools regularly, prototype your own ideas, and can say specifically what you would change about a model's behavior and why. You know the trade-offs that matter: context windows, latency, cost, hallucination, and RAG versus fine-tuning.
You have built evals (golden sets, ground truth, and offline and live evals) and used precision- and recall-based metrics to make product decisions, and you have set eval standards that other teams adopt.
When you find a problem, you build the infrastructure that prevents the whole class of problem, and you make that infrastructure reusable across the organization.
You move fluently between business and technical concepts, with a track record of business cases, ROI models, and roadmaps that inform investment decisions at the executive level.
A deep working understanding of modern AI/ML and GenAI (LLMs, agents, RAG, prompt/harness engineering, and Evals) and how it applies to enterprise problems.
You work from first principles, prioritize well, and make good calls quickly with incomplete information.

Nice To Haves

12+ years of product experience, or an equivalent depth of AI-native building, in a large enterprise environment.
Experience in healthcare, insurance, or another regulated industry.
A well-supported, publicly demonstrated point of view on where agentic AI, evals, and AI-native product development are headed.
Hands-on experience with agent frameworks, GraphRAG / knowledge graphs, and reusable skills / plugins.
Experience defining a new AI product category or standing up cross-organization AI standards and playbooks.
Familiarity with responsible AI, bias mitigation, and compliance with sensitive data.
A degree in computer science or engineering, or an equivalent combination of experience.

Responsibilities

Define the multi-year vision and strategy for a portfolio of AI products and capabilities. Prioritize across competing product areas and adapt the portfolio as the model landscape and the business change. Set the north-star and the long-term roadmap, and reason from first principles when the problem, opportunity, and strategy are not yet defined.
Partner with business-unit and enterprise leaders to map their operations end to end. Turn ambiguous, high-stakes problems into a slate of AI use cases, each with a clear business case, and sequence them into a portfolio.
Build working prototypes and demos with agentic coding tools (Claude Code, Cursor, OpenAI Codex) to prove feasibility and value before a full team is committed. Stay close enough to the technology to make credible technical calls.
Turn validated use cases into buildable requirements: user stories, acceptance criteria, edge cases, and success metrics. Shape the end-to-end user experience with design, because adoption depends on usability.
Drive the build from prototype to production across your portfolio. Sequence the work, make scope and tradeoff calls as models and constraints shift, unblock teams, and keep momentum from first demo through launch and iteration.
Define how quality is measured. Build golden sets, error budgets, and eval frameworks; run offline and production evals; choose the right metrics (precision, recall, task success, hallucination rate); apply LLM-as-judge where it has been validated. Turn results into launch decisions and leadership readouts.
Partner with engineering on high-level architecture and feasibility across the portfolio, while owning the context engineering and skill-creation components of the AI harness and setting the reusable patterns others build on.
Lead engineers, data scientists, designers, product managers, and business partners without direct authority. Influence VP-level organizations through written narratives, business cases, and executive forums. Author the playbooks, skills, and solution patterns for building AI at Humana.
Move solutions through Humana's enterprise AI governance (AIRB, LRC, Responsible AI Council), owning the product documentation, scorecards, and stage-gate reviews, and help shape the governance standards themselves as the portfolio grows.
Raise the bar for the whole team. Mentor senior and lead product managers, contribute to hiring and promotion assessments, and set the craft standard for building with AI across the organization.
Communicate strategy, roadmap, requirements, ROI, and eval results clearly to executive, business, and technical audiences, and represent Humana's AI product work externally with partners, vendors, and the broader community.
Track advances in models, agents, evals, and emerging techniques (agent harness, loop, and graph / GraphRAG approaches), form a defensible point of view on where they are headed, and apply them across your portfolio.