Soon
Before AI can answer, someone has to put the knowledge in order
We're working on Ragen Brain
Most companies don't have an AI problem. They have a scattered-information problem: documentation in three places, decisions buried in email, a procedure that exists in two versions and nobody knows which one is current.
Feeding that into an assistant doesn't clean it up. It gives you faster access to the same scattered information.
Ragen Brain is the part of Ragen that handles what happens before the question. It analyzes documents, notes and scattered decisions, and gives back knowledge you can actually read: a list of topics, files, and a graph of how people, processes, customers and products connect.
Along the way it surfaces what usually nobody sees:
what contradicts itself
two documents saying different things about the same thing
what's missing
topics everyone knows about and nobody wrote down
what's out of date
knowledge nobody has touched in a long time, with no owner
Three rules we're building on
Nothing enters the knowledge base without a human approving it.
Brain proposes, someone at your company approves. Software that decides on its own what's true about your company isn't a feature, it's a risk.
You set permissions once, while you're sorting things out.
Not per file, but per piece of information, together with the person who owns it. Ragen respects that on every answer afterwards.
The result is yours, and readable without us.
Organized knowledge is plain files you can open in any editor and version like code. Importing them into Ragen takes one move, but you don't have to. If you stop at the cleanup, that's a result too.