Principal Product Manager, Agent-First

What you'll do

• Own the customer-facing AI experience end-to-end. Define the core flows, ship them, instrument them, iterate. The quality of these flows is the product.
• Make the AI feel trustworthy. Small business owners won't hand over real work to an AI that doesn't show its reasoning, respect their constraints, or earn trust gradually. You'll define how the product communicates intent, surfaces tradeoffs, and graduates customers from supervised to autonomous use.
• Pick the right leading indicators. Activation, time-to-value, weekly engagement, owner intervention rate, attributed outcomes. Retention is the lagging story — these are the levers. You'll commit to numbers and defend them.
• Partner across product surfaces. Our customer-facing experience sits on top of shared platform capabilities used by other parts of the business. Coordinate with the PMs and engineers who own those layers to ship fast without creating tax for either side.
• Be opinionated about AI UX. Decide where LLMs make the product magical and where they make it brittle. Own the eval bar for agent behavior — accuracy, safety, brand tone, customer-appropriateness. Have a strong point of view on when not to use AI.
• Use AI to build, not just to ship. Make AI-assisted research, prototyping, and iteration default working modes for the team. Move from idea to tested artifact in days, not sprints.
• Get out of the building. Talk to small business owners weekly. Sit in support tickets. Bring their voice into every roadmap conversation, loudly.
• Communicate up and across. Pitch the roadmap to executives, jam with engineering on architecture, defend tradeoffs to design, and translate customer pain into requirements that engineers can actually act on.

Who you are

• 8+ years in product, with at least one specific example of 0-to-1 product you took from concept to scale. SaaS experience required; SMB experience strongly preferred.
• Genuinely AI-fluent. You've shipped LLM-powered features in production. You understand prompt engineering, RAG, evals, and agent architectures well enough to argue with engineers about tradeoffs, and not just hand them a PRD.
• Direct experience building agentic systems (not just AI features bolted onto existing products).
• Sharp on AI UX.