Lore is knowledge capture for AI agents. It interviews employees, turns the unwritten rules of their work into sourced rules a manager approves, and serves those rules to your AI agents through an API.
You add the person in Lore and send them a private interview link. Lore chats with them in plain text for about 30 to 45 minutes and keeps asking about exceptions, approvals, clients and workarounds until the specifics come out. They can pause and come back any time.
When you add someone, choose whether they are leaving or staying. For people who are staying, the interview captures what they know so colleagues and agents can rely on it without always asking them.
Your agent sends a question to one endpoint, POST /api/v1/query, with an API key. It gets back an answer plus the approved rules behind it and who each rule came from. You can also export every approved rule as JSON or Markdown and load it into the tools you already run.
A wiki only holds what someone wrote down. Lore goes after the exceptions people never write down, and every rule keeps the condition, the action, the reason and the exact quote it came from, so your team can check it.
Hi Product Hunt, I'm Numair, 18, and I'm building Lore.
Every company I talked to had the same problem. They're rolling out AI agents, the agents read the wiki, and the wiki is missing the part that matters: the client who needs a credit before the ticket closes, the approval only one person gives, the workaround nobody wrote down. When that person leaves, it's gone.
Lore interviews people in a short text chat and keeps asking "when isn't it usual?" until the specifics come out. Every answer becomes a rule with a condition, an action, a reason and the exact quote. A manager approves each one before any agent sees it, and if Lore doesn't know something, it says so instead of guessing.
Sign up with a work email and you get a free trial with 2 interviews. I'd love to hear one thing from you: what's a rule at your company that only one person knows?