Mnemocompany memory

Manifestus

Our mission

Most of how a company works was never written down. Why refunds over a certain amount need a second sign-off. Why deploys pause on Fridays. Which account got a custom SLA, and who agreed to it. It lives in Slack threads, ticket comments, and the heads of the people who were there.

Then someone changes roles, or leaves, and it walks out the door with them. The next person finds a stale wiki page, or a search result for a decision that was reversed months ago, and can't tell which one still holds. AI agents hit the same wall, just faster. They'll follow last quarter's rule without blinking.

We're building Mnemo so that knowledge stays. It learns from the work your team already does in Slack, email, Jira, and GitHub, and keeps the decisions, exceptions, and lessons behind it. When a newer decision replaces an older one, the old one is marked superseded and ranks lower. It isn't deleted, so you can still ask what changed and when. Answers come back with the message, ticket, or doc they came from.

People ask in Slack. Agents ask through the API or MCP. It's the same memory either way. We're starting with handovers, because that's where the most gets lost: the next person should be able to ask why things are set up the way they are, and get the thread where it was decided.

We show our work

Memory you can't check isn't worth much, and that goes for our own claims too. We don't hide behind proprietary scores. We publish our results, our methodology, and the code to reproduce them.

See the benchmarks and how to rerun them: https://mnemohq.com/benchmarks#methodology

Explore Mnemo