Technical Product Manager

Keystone AI is building the self-maintaining knowledge base — continuously monitoring, fixing, and serving trusted context to every human, agent, and system that needs it. The context crisis is real: agents now act on stale docs at scale. You're joining at the moment where we build our biggest evolution to close this loop. This is a founding product role in the agentic era, and you'll shape how knowledge stays alive. You'll be the only PM on the team, working closely with Chris (CEO) and Pierre (CTO) and embedded in a ~10-person engineering team. This is not just a coordination role; you're a builder. Your job: meaningfully expand the team's output, co-own the roadmap, and make the hard calls on scope. You'll be in the room for major decisions, and you'll be accountable for them. What you'll do Co-build a strategic roadmap that proves the new self-maintaining value prop of Keystone AI , and help us close large deals and grow our ACV. Define, challenge, and cut scope with conviction. You own what gets built and what doesn't. Work daily with engineers on hard, ambiguous problems (AI, knowledge drift detection, doc verification). Own the AI product surface — speak fluently with engineers about LLMs, embeddings, agents, and drift detection. You'll shape how Keystone AI Agent closes the loop. Talk to users, read support threads, and bring real-world signal into engineering conversations. Identify what we should stop building and what we should start, and make the case for both. Tech You have seniority. You've shipped products that matter at scale. You're comfortable in leadership conversations, contributing to decisions and reshaping the roadmap. This is not a first PM job. Engineer past. You can code. Not "you've worked with engineers": you've written production code. You can read a PR, have an opinion, and contribute more now that you have AI in the loop. Strong IC, and you enjoy it. We're a team of 20, so this isn't a classic PM role; it comes with a meaningful individual contributor component. AI fluency. You speak AI natively. LLMs, embeddings, agents, RAG, drift detection: not buzzwords, your daily vocabulary. You could hold a deep technical conversation with our CTO without needing a translator. Cultural fit. You thrive in a team of ~20 people across 8 countries with no hand-holding. You're naturally curious, you move fast, and you own outcomes. You don't wait for process; you create it. Originally posted on Himalayas