Staff Product Manager, AI Foundations

Vanta is a trust management platform that helps businesses continuously monitor, improve, and prove their security and compliance. The Staff Product Manager, AI Foundations will own the strategy and roadmap for the agent harness and AI quality and evaluation stack, building shared capabilities that improve the quality, safety, reliability, and trustworthiness of Vanta’s AI products.

Responsibilities:

• Own the strategy and roadmap for Vanta’s core AI capabilities: the agent harness and the AI quality and evaluation stack
• Live in agent traces to build intuition for how customers actually use our AI products, and turn that into the highest-leverage improvements
• Make Vanta’s AI products win on quality: build golden datasets, automated evaluators aligned to expert judgment, and self-serve tooling so PMs, subject-matter experts, and designers can run evals without an engineer in the loop
• Scale the agent harness: shape the core capabilities Vanta’s agents are built on, making clear tradeoffs across cost, latency, reliability, and safety
• Set the quality bar teams adopt for evals, launch readiness, and observability, and drive adoption without becoming a gatekeeper
• Partner closely with EPD teams across Vanta, and with the AI pillar’s other teams, to turn proven patterns into shared capabilities
• Help define and shape Vanta’s product team culture as we grow

Requirements:

• Have 8+ years of product management experience, including platform, infrastructure, or technically deep products
• Have hands-on experience building with LLMs and AI agents, and understand how evals, observability, and quality signals drive systematic improvement in AI products
• Bring strong user empathy: you get close to customer problems and bring that clarity back to the team
• Know how to run foundational products like a product, with adoption goals, feedback loops, and prioritization grounded in what your users need
• Earn credibility with senior engineers and hold your own in deeply technical design discussions
• Use coding agents daily, and stay deeply curious about where models and agent tooling are going
• Think in systems: you can reason about how evaluation, observability, and infrastructure choices compound across many teams and customer experiences
• Thrive in ambiguity in a fast-moving AI space, leading with the problem and the process rather than a fixed answer
• Communicate with clarity and influence, both written and verbal, with team members, stakeholders, and executives
• Open to using AI to amplify their skills and strengthen their work - demonstrating curiosity, a willingness to learn, and sound judgment in applying AI responsibly to improve efficiency and impact