Product Manager II, Agent Development

Whoop • Boston, MA

2d • $125,000 - $170,000 • Onsite

About The Position

At WHOOP, we’re on a mission to unlock human performance and extend healthspan. We empower our members with real-time, longitudinal health insights that help them understand how their daily choices compound over time—enabling better long-term health and lasting behavior change.

As a Product Manager focused on Agent Development at WHOOP, you will partner closely with Product Managers, AI Engineers, Performance Science experts and Data Scientists to design, write, and continuously improve AI agents that power health coaching experiences across the app.

You will play a central role in shaping how WHOOP AI systems are designed—how agents understand members, reason over their health data, coordinate with one another, and communicate insights in a way that is accurate, empathetic, actionable, and behavior-changing.

Requirements

• 3+ years in product management or related technical or strategic role, with at least 1+ year working on LLM-focused or agentic AI applications
• Deep understanding of LLM engineering concepts (prompt engineering, RAG, tool usage, real-time decisioning)
• Comfortable with Ambiguity: Proven ability to define product requirements and success metrics in ambiguous environments
• Exceptional written and language design skills, with strong instincts for clarity, tone, and structure—and a deep understanding of how language influences trust, motivation, and behavior change in sensitive health contexts
• Experience designing high-quality conversational AI, including prompts, system instructions, coaching narratives, and microcopy, with the ability to treat language as a control surface for agent behavior, not just user-facing copy
• Strong product and technical fluency, enabling deep collaboration with AI Engineers and Data Scientists and informed reasoning about LLM behavior, limitations, failure modes, and system tradeoffs (without requiring production ML development)
• Systems-level thinking, with experience translating ambiguous problems into well-scoped agent behaviors, modular workflows, and clear specifications that scale beyond a single agent or feature
• Analytical and strategic judgment, including the ability to diagnose agent and system failures, define quality metrics and evals, and balance short-term execution with long-term platform, scalability, and quality considerations
• Relevant technical or interdisciplinary background, such as Computer Science, Engineering, Mathematics, Cognitive Science, HCI, Linguistics, or equivalent practical experience at the intersection of technology, product, and customer experience (MBA a plus)

Responsibilities

• Design and write AI agents that help members understand their data, set goals, and make better daily decisions about sleep, recovery, training, and long-term healthspan
• Design complex, multi-agent systems where specialized agents (e.g., data analysis, health reasoning, coaching narrative, safety/guardrails) collaborate to produce a single coherent member experience
• Partner with AI Engineers to shape agent orchestration, including sequencing, handoff, tool usage, memory access, and fallback behaviors
• Own the conversational and behavioral strategy for agents—defining tone, structure, narrative flow, and language quality across proactive reports and interactive coaching
• Translate ambiguous cross-functional needs into well-scoped agent capabilities, turning inputs like “members don’t understand recovery” or “support is overloaded” into clearly defined agent roles, inputs, outputs, and metrics
• Shape how data is prepared and presented for agent reasoning, including defining the right aggregates, comparisons, and summaries needed for high-quality outputs
• Write and maintain evaluation frameworks (evals) to systematically measure and improve agent quality at both the individual-agent and end-to-end system level
• Define what “good” looks like for each agent and system, including success criteria, failure modes, and quality bars across correctness, personalization, actionability, and trust
• Use qualitative and quantitative feedback (member conversations, internal beta feedback, eval results, engagement metrics) to inform agent improvements and product direction
• Act as a trusted advisor to teams across WHOOP on AI strategy, agent design, and responsible use of generative AI in health contexts

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level
• Education Level: No Education Listed