AI Product Lead (Senior Product Manager) at EverQuote in Cambridge HQ

Job description

Location Remote About Us EverQuote is a leading AI-powered growth solutions partner for regulated property and casualty insurance entities, enabling the largest insurance carriers and thousands of agents to maximize customer acquisition across digital channels. Fueled by our proprietary data assets and our AI traffic engine, EverQuote is transforming the way providers attract and engage consumers to grow market share. Joining our team puts you at the forefront of this transformation. You will collaborate with a group of builders to solve complex problems and reshape a massive industry. If you want to drive real-world impact and do the best work of your career, you’ve found your home. We move fast, follow the data, and hold ourselves to high standards. If that's how you work too, this is the role for you. About the Role As the AI Product Lead you will own a product unlike anything we have built: an insurance marketplace that AI agents use to shop on a person’s behalf. When an AI agent shops for insurance, you will make the AI- insurance shopping experience the one it trusts and can buy through. You will help shape what agentic insurance shopping looks like before the market arrives. You own the agent-facing product: the data an agent sends and gets back, the recommendation and reasoning we return, the MCP server and skill that let AI platforms connect to us, and the trust signals that make an agent willing to rely on us. You own the roadmap, the testing of how well the agent performs, and the relationships with the AI platforms whose agents call us. Often your customer is another AI agent rather than a person, and the product has to earn that agent’s trust What You’ll Do Define and own the product roadmap for the marketplace, balancing what agents need, what the architecture allows, and where the business is going. Own testing and quality: build and maintain the eval sets, name the ways the agent can go wrong, and decide what counts as a good agent interaction. Track how the agent is doing: completion, accuracy, speed, and how often a handoff is followed through. Build quick prototypes while the problem is still fuzzy: stand up agent flows and put them in front of real and test agents to learn before engineering commits. Take what we learn about how agents shop and turn it into real changes in the data model, the reasoning we expose, and the architecture. Work with the AI technical lead and engineers on architecture trade-offs, cost, and speed, and keep the shopper’s and the calling agent’s interests in every decision. Build relationships with AI platforms to learn how they find and route to agents, and bring that back into our design. Work in the open. Help publish our mission, experiments, and findings. Give leadership clear, regular visibility into progress, what we are learning, and risks. What You Bring to the Team (Qualifications) Experience building and shipping real AI products, ideally from an AI lab, an AI-native company, or a team that has put LLM or agent products into production. You have done this, not just studied it. Comfort with testing AI: you can take a fuzzy goal and walk through how you would test it, where it would fail, and when you would roll it back, before reaching for a model. A working understanding of agent systems: LLMs, how agents are put together, retrieval, orchestration, guardrails, and the cost and speed trade-offs that come with them. You do not need to build the models, but you need to talk about them with engineers. Enough technical depth to hold your own in an architecture or code review and ask a sharp question. Comfort with light Python or SQL to look at data and prototype. Strong product instincts and comfort with ambiguity. You can make progress on loosely defined problems. Clear communication with both technical and non-technical people, and the habit of leading with data. A continuous learner, with grit and a bias for action in a space where the standards change month to month. 8+ ye