Your memory should be yours.
Switch AIs and you start over. You explain who you are, what you're working on, and how you like things done, then you do it again in the next tool, and the next. When an AI does remember you, that memory is locked inside it, owned by whoever built it, and it doesn't follow you anywhere else.
We think that's backwards. The memory you build by talking to AI is yours. It should follow you everywhere, and you should decide what goes in it.
You choose what your AI remembers.
Tabula doesn't quietly record everything you say. It keeps what you or your AI decide is worth keeping. A second brain full of noise is just noise. Yours stays deliberate, and it stays yours to read, edit, or delete whenever you want.
Your memory should work in every AI, not one.
Save something in one AI, use it in another. Tabula runs on MCP, the open standard that every major AI already speaks and that no single company owns. We don't have to talk the industry into connecting. It already has. Every new AI that adopts the standard is one more place your memory works, with nothing for you to do.
Your memories aren't a product.
We don't train on your memories. We don't sell them. We never hand them to outside AI companies. They're yours, and they stay that way.
This is early, and we won't pretend otherwise.
Connecting an AI still takes a few minutes, once per tool. You don't need to be a developer, but you do have to set it up, and we won't sell that as magic. MCP, the open standard Tabula runs on, is young: when we started only ChatGPT, Claude, and Mistral spoke it in chat, then Grok came, and more will follow. Every release makes connecting simpler and recall more automatic. We're building for the day your memory just follows you with nothing to set up, and we'd rather be early and honest about the rough edges than wait and sell you a promise.
One memory that's yours, in every AI you use.