Lead Product Manager, Agentic AI

Wolters Kluwer•Kennesaw, GA

1d•$107,500 - $188,400•Onsite

About The Position

We are seeking a Lead Product Manager for Agentic AI to lead a small, highly autonomous pod focused on building a new agentic AI product within our Tax & Accounting division. This role requires a blend of product strategy, tax domain logic, and AI capability. The Product Manager will be the central figure for AI execution, working closely with domain experts to understand tax logic and using agentic tooling to prototype workflows for engineering. A key aspect of this role is orchestrating agents around existing tax calculation engines rather than reimplementing deterministic logic within prompts.

Requirements

Daily practitioner of agentic tooling, with experience orchestrating work using environments like Claude Code, Cursor, or Copilot.
Ability to debug agent failures by analyzing raw transcripts to identify root causes (prompt issues, missing tools, eval gaps, model limitations).
Systems thinking, including understanding of context windows, RAG, model routing, tool design, and when to use deterministic rules or existing engines over LLM calls.
Proven track record of shipping user experiences that honestly handle AI uncertainty, including fallbacks, confidence signals, and human-in-the-loop validation.
Commercial pragmatism with a bias for time-to-market, including knowing when to use third-party APIs versus investing in proprietary orchestration.
Experience shipping commercial software in regulated enterprise SaaS environments where compliance, data privacy, auditability, and uptime are critical.
Ability to identify gaps in team capabilities and help build the right team around the product.

Nice To Haves

Experience in tax, accounting, legal, healthcare, or financial services.
Ability to appear onsite at a Wolters Kluwer office as part of the recruitment process.

Responsibilities

Develop and evaluate high-fidelity agent prototypes and their corresponding eval suites, which will serve as the product specification.
Own and design the eval framework to measure accuracy, detect hallucinations, and identify logic failures in tax software.
Translate complex tax jobs-to-be-done into structured agent architectures, including multi-agent orchestration, state management, and human-in-the-loop decision points.
Manage the unit economics of AI models, balancing accuracy with latency and inference costs, and designing cost-effective alternatives.
Design systems for graceful degradation, including explicit fallbacks to deterministic rules or human intervention when agents encounter ambiguity.
Partner closely with Tax Subject Matter Experts (SMEs) to translate their judgment into system prompts, tool contracts, and eval criteria, building trust and collaborative relationships.
Oversee the product journey from opportunity identification through validated prototype, evaluation performance, and production launch, with success measured by customer adoption, accuracy, and business impact.