AI Product Manager - Agency Exp Required

We're looking for a Digital Product Manager who can own a client engagement end-to-end — from scoping to shipped product — across three kinds of work: building new AI agents, standing up greenfield products the agentic way, and helping established companies rebuild aging platforms and level up their teams to work AI-natively. You'll sit between clients and agentic engineering teams, turning ambiguous business problems into scoped, sequenced, shippable work — whatever form that work takes.

This isn't a role for someone who's only managed roadmaps in a single-product SaaS company, and it isn't a narrow "AI agent PM" role either. Some engagements will be zero-to-one greenfield builds. Some will be rebuilding a client's legacy platform without breaking what already works. Some will involve teaching a client's own team how to build and ship AI-natively so they don't need us in the room forever. You've worked in an agency or consulting environment, you know how to run multiple client relationships at once, and you know how to say "no" to a client gracefully when the ask doesn't hold up.

• Own client engagements end-to-end — from discovery and scoping through delivery, acting as the primary product voice with the client, not hidden behind an account manager.
• Scope and run greenfield product builds — take a zero-to-one idea and turn it into a sequenced, shippable roadmap built the agentic way from day one.
• Lead platform rebuild and modernization engagements — help clients replatform or rebuild legacy systems without stalling their existing business, balancing technical debt against delivery speed.
• Design and help deliver AI-native enablement — work with clients' own product and engineering teams to build their internal muscle for agentic development, not just hand off a finished product.
• Translate ambiguous client asks into scoped work — write PRDs, define success criteria, and sequence delivery so engineers know exactly what to build and why.
• Partner directly with agentic engineers — understand the difference between a deterministic workflow and a true agentic use case, and help the team pick the right one for each engagement.
• Set guardrails and evaluation criteria for AI agents before they touch a client's real data or production environment.
• Manage multiple concurrent engagements — prioritize ruthlessly across clients with competing timelines, budgets, and engagement types.
• Run the room — client steering committees, sprint reviews, and the hard conversations when scope, timeline, or feasibility shifts.
• Report on outcomes, not just output — define what success and ROI look like for each engagement type, and hold the team accountable to it.
• Bring craft to how TribalScale delivers AI-first work — contribute back to internal playbooks on scoping agentic builds, platform rebuilds, and client enablement.

• 5+ years in product management, with real ownership of scope, roadmap, and delivery outcomes.
• Agency or consulting experience — you've worked across multiple clients or accounts, not just one internal product team. You know how to manage client expectations and protect scope.
• Range across engagement types — you've run at least two of: a zero-to-one greenfield build, a legacy platform rebuild/migration, and a client-facing enablement or training engagement.
• Hands-on experience with AI agents — you've scoped, built, or shipped agentic products or workflows (not just prompted a chatbot). You understand the tradeoffs between agentic and deterministic approaches.
• Comfort working alongside agentic engineers — you can hold a technical conversation about architecture, evals, and guardrails without needing everything translated for you.
• Teaching instinct — you can explain agentic concepts to a client's non-technical stakeholders and their engineers, and you get satisfaction from leveling up someone else's team.
• Strong client-facing presence — you can present to a VP-level stakeholder, handle pushback, and say no when it's warranted.
• Sharp scoping and prioritization instincts — you can turn a vague brief into a one-page plan quickly and defend the tradeoffs.
• Comfortable in ambiguity — you don't wait for a fully-formed spec to start moving.

• Experience with LLM evaluation frameworks or agent observability tooling.
• Background in a technical role (engineering, data, or QA) before moving into product.
• Experience managing engagements in regulated industries (finance, healthcare).
• Experience designing or delivering technical training, workshops, or enablement programs for client teams.
• Track record of writing or speaking publicly about product or AI topics.