AI Product Manager

Company and product

Role overview

Product strategy and discovery

• Define AI/ML product vision, roadmaps, success metrics, and high-impact use cases.
• Translate business problems into AI capabilities and system designs; balance feasibility, GPU and infrastructure costs, performance, and ROI.
• Research user workflows and friction; validate ideas through experiments, pilots, and proofs of concept.
• Define measurable hypotheses involving latency, accuracy, cost per request, and adoption.

AI system delivery and quality

• Partner with engineering on RAG architectures, agentic workflows, model selection and benchmarking, prompt strategies, evaluations, and guardrails.
• Write PRDs and technical specifications; prioritize features by impact, risk, and complexity; lead sprint planning and cross-functional execution.
• Define AI quality metrics, implement evaluation pipelines and red-teaming, and drive continuous improvements through experiments and feedback.

Governance and go-to-market

• Address data privacy, bias mitigation, model governance, and compliance; partner with security and legal teams on guardrails and safe deployment.
• Align leadership, engineering, and go-to-market teams; communicate technical tradeoffs to non-technical stakeholders.
• Support launches, positioning, and customer enablement; track adoption and business outcomes after launch.

Required skills and experience

• 4–7+ years in Product Management or technical product roles.
• Proven experience shipping complex software or platform products and owning roadmaps and PRDs.
• Knowledge of LLMs and foundation models, RAG pipelines, agents and orchestration frameworks, APIs, and system architecture.
• Ability to read architecture diagrams and collaborate deeply with engineers.
• Strong prioritization and tradeoff decision-making, data-driven thinking, comfort with ambiguity and rapid iteration, and excellent written and verbal communication.

Additional areas of familiarity

• Familiarity with prompt engineering, model evaluation and benchmarking, cost optimization for tokens, GPU, or infrastructure, and experiment design such as A/B testing or offline evaluations.