Eighteen months ago we launched our first AI feature. Customers loved the concept. The UX was confusing, the model made too many visible errors, and support tickets tripled in the first week. We pulled it, spent four months rebuilding the product feedback loop, and relaunched. It now drives 28% of our expansion revenue and is the most-mentioned feature in NPS surveys. We learned more from that failure than from any success. We are now hiring an AI Product Manager to own the next generation of AI features — from writing the product requirements document through working with our ML team on model tradeoffs to managing expectations with sales and customers. You don't need to know how to train a model. You do need enough intuition about AI systems to know what's feasible, what's risky, and what questions to ask before committing to a roadmap.
Responsibilities
Own the AI feature roadmap and communicate it clearly to all stakeholders
Write product requirements documents with clear success metrics and edge case coverage
Work with ML engineers on model tradeoffs and feasibility assessments
Run monthly user research sessions and synthesise findings into product decisions
Define and monitor KPIs for every AI feature in production
Requirements
3–5 years of product management experience with at least one AI feature shipped
Able to translate between business requirements, user needs, and ML team constraints
Comfort with product analytics — SQL queries, funnel analysis, A/B test interpretation
Strong written communication — you write PRDs that engineers actually want to read
Experience working in an async-first, distributed team
Familiarity with OpenAI or similar APIs enough to write accurate acceptance criteria
Benefits
Real ownership — your decisions ship to thousands of users