Senior Product Manager, AI

ServiceUp•Campbell, CA

3d•Remote

About The Position

ServiceUp is reshaping how fleets and shops coordinate vehicle repair, and AI sits at the core of how we scale that operation. We are hiring a Senior PM, AI to own the end-to-end experience our fleet managers, shops, service advisors, and drivers have with the platform — both the agentic surfaces they interact with every day and the integrations that connect ServiceUp to the systems they already run their businesses on. This is a model-native, evaluation-driven product role at the heart of a multi-sided marketplace. You will define what "good" means for an LLM-powered workflow, build the eval systems that prove it, and iterate at a velocity unique to AI products — while also driving the integrations and APIs that make ServiceUp the connective tissue between fleets, shops, and the broader repair ecosystem. You will partner closely with engineering, applied AI, design, and operations to ship customer-facing agents and the platform that powers them.

Requirements

10+ years of product management experience, with at least 2 years shipping production AI or agentic products to real users (internal or external), and at least 5 years building or scaling a marketplace.
Multi-sided marketplace fluency. Deep understanding of how marketplaces actually work — supply and demand dynamics, liquidity, network effects, take rate — and how to balance the often-competing needs of distinct user types on the same platform (e.g., fleets vs. shops, buyers vs. operators).
APIs & integrations in your DNA. You read API docs for fun, can scope a partner integration end-to-end, and have shipped products where the integration is the product. Strong instincts on REST, auth flows, data contracts, partner enablement, and the operational realities of keeping integrations healthy at scale.
Hands-on AI fluency. You can read a prompt, reason about tool calls and context windows, and form a real opinion on model choice, retrieval strategy, and agent architecture. You understand the difference between a demo and a system.
Systems thinking for agents. You think in workflows, state, and failure modes — not just screens. You know when to use a deterministic pipeline, when to give a model agency, and how to design the seam between them.
Design partnership with modern tools. Strong taste and a track record of partnering closely with designers — including using modern AI design tools (Figma AI, v0, generative UI, rapid prototyping) to explore experiences and pressure-test ideas before engineering investment.
Strong product fundamentals. User research, scoping, written specs, prioritization, instrumentation, and shipping. AI does not replace the basics; it raises the bar on them.
Quantitative rigor. Comfort writing SQL, exploring product data, and defining the metrics — including model and agent quality metrics — that govern the roadmap.
Operator empathy. Curiosity for the messy reality of fleet ops and auto repair. You will talk to service advisors, shop owners, and fleet managers regularly, and you will enjoy it.

Responsibilities

Customer-facing product — agentic and integrated. Define and ship AI experiences for fleet managers, shops, service advisors, and drivers alongside the integrations that make those experiences feel native inside the tools each side already uses.
Marketplace integrations & APIs. Own the integration surface between ServiceUp, fleet platforms, and shop systems. Define API contracts, partner onboarding flows, and data models that let both sides of the marketplace plug in quickly and reliably.
AI platform & tooling. Drive the internal platform that powers every agent at ServiceUp: prompt and tool registries, agent orchestration, model routing, retrieval, observability, guardrails, and human-in-the-loop review. Make it easy for product teams to ship safely and quickly.
Evaluation as a product surface. Treat evals as a first-class deliverable. Design offline and online eval suites, golden datasets, LLM-as-judge rubrics, and live shadow-mode tests. Decide what to ship based on measurable agent quality, not vibes.
Trust, safety, and reliability. Define escalation paths, fallback behaviors, confidence thresholds, and review workflows so customers can trust agent and integration output for high-stakes decisions like estimates, approvals, and dispatching.
Experience design partnership. Work hand-in-hand with designers — leveraging modern AI-native design tools (Figma AI, generative prototyping, rapid mock-up systems) — to compress the loop from idea to testable experience and to ship interfaces that feel effortless for non-technical operators on every side of the marketplace.
Cross-functional leadership. Partner with engineering, design, ops, and GTM. Translate ambiguous customer problems into crisp specs, technical bets, and shippable increments. Bring senior judgment to scope, sequencing, and trade-offs.