Senior Product Manager, AI Agents
at WorkHero · 11-50 employees
- Seniority
- Senior
- Work model
- Remote
- Employment
- Full Time
- Location
- USA (Remote)
- Posted
- 5d ago
at WorkHero · 11-50 employees
WorkHero provides an AI-powered back-office platform that autonomously manages administrative tasks for small HVAC contractors.
WorkHero is building the AI-powered back office for the skilled trades, starting with the $50B+ HVAC industry. Small contractors are great at their trade but lose 20+ hours a week to invoicing, permits, and paperwork. WorkHero has a team of expert office managers who run their back office, and their work is backed by automation and AI tooling that lets a small team do far more than they otherwise could. We launched in late 2024 and work with contractors across the country. We're hiring a senior product manager to build the AI agents at the core of the product. It's a hard problem: getting an agent to reliably run a business's back office. It's also what determines how fast we grow, because the stronger these agents become, the more contractors each office manager can serve. This role is a key part of our company's next stage. You'd own that work alongside a small, senior team with deep experience in AI, product, and the trades. ABOUT WORKHERO Tens of thousands of small HVAC contractors, who keep our homes comfortable and enable the climate transition, are drowning in administrative work. Legacy software is clunky, full-time admins are expensive, and owners spend 20+ hours per week on paperwork. WorkHero changes that. Our platform combines agentic AI with human-in-the-loop operations to autonomously handle back-office workflows like invoicing, permits, rebates, pricebooks, warranty registration and a whole lot more. Instead of asking contractors to learn new tools, our AI agents do the work, with expert office managers providing oversight and operational coaching. The result: real work gets done autonomously, helping small contractors grow profitable, resilient businesses while accelerating heat pump adoption nationwide. We've raised over $8M to date and are backed by Navitas Capital, Workshop Ventures, York IE, and strategic angels including a former ServiceTitan executive. Our advisors include Chad Laurans (Founder of SimpliSafe) and Gerald Chertavian (Founder of Year Up). THE ROLE You'll be an early product hire working alongside our founding PM and co-founder. Where they hold strategy across the whole product, you go deep on the engine that does the work: our AI agents. You own the product question of whether each is genuinely accomplishing the user's goal, and how we know. This is a senior individual-contributor role built around depth. You'll own one or two agent experiences end to end, working close with engineering, service, sales, and the customer. You'll report to our founding PM, who holds product strategy and prioritization across the whole product. Within your area, you own the roadmap, the craft, and the results, and the closer you get to users, the more your work shapes where we go next. We're looking for a high-craft PM who wants to be in the weeds on hard problems: shaping how an agent's prompt, tools, and knowledge cohere into one experience, holding a high bar on quality, and building the evals and instrumentation that show whether it works. Crucially, this is product work: you keep the user's goal at the center and treat evals as the instrument, not the job. This is an AI-native role. You'll shape agentic systems, and you'll lean on AI heavily in your own work. If going deep while keeping the bigger picture in your head sounds like your kind of product work, keep reading. WHAT YOU'LL OWN Own your workstream end to end. Own and evolve the roadmap for your product area: dependencies, delivery, and the calls to make when priorities shift. Take features from requirements through launch, documentation, and iteration, and keep a clear line of sight to whether they move adoption, retention, and revenue. Turn ambiguity into product. Translate messy, ambiguous user needs into concrete AI behaviors, workflows, and requirements. Boil complex technical tradeoffs down into product decisions your team can act on. Stay close to users. Spend real time with our Senior Office Managers and contractor customers. Run discovery that gets past the surface: ask "why" with empathy, listen hard, and let what you learn drive what you build. Define with engineering. Own the use cases, specs, and tradeoffs for your area. You're in the weeds by preference, not throwing specs over the wall, and you bring real UX instincts for the unstructured, AI-driven workflows this product runs on. Hold the bar. You have a clear sense of what good looks like, and you keep raising the quality bar in your area. QA here is a partnership with our product-minded engineers, not a gate you stand at, and you help the team ship work everyone's proud of without slowing work down. Prove it's working. Define success metrics across customer satisfaction, operational efficiency, and agent performance, and drive continuous improvement through data, user feedback, and experimentation. Build the measurement muscle we don't have yet: own the KPIs, stand up the evals and instrumentation, and feed what you learn into the next round. WHO WE'RE LOOKING FOR You have strong product management experience. You've owned products from strategy through launch, and you move easily between big-picture thinking and hands-on execution. You have real product judgment: you make the hard calls on priorities and tradeoffs, you know what to say no to, and you back your decisions with the right amount of data and research. You bring 4–6 years of product management experience. You go deep while holding the big picture. You can live inside the details of a workstream and still hold the wider product and business in your head. You know which details matter and which don't. You think in AI. You're curious and creative about what's newly possible, and happy to rethink a user experience from first principles rather than port over how it worked before. You understand LLMs well enough to reason about their limits and tradeoffs, and you can test an LLM pipeline in a structured way using no-code and prototyping tools. You use