One brain per client, no mixing between projects.
What is known about each project, gathered and indexed in one place and isolated per client. The agent of one project cannot see another's notes.
The problem
“The person who ran that project knows, and they are out today.” “It is in some ticket, or in a document, I cannot remember which.” “What if the AI mixes one client's material with another's?”
It happens to any company running several projects or several clients at once: what is known is scattered, and handing it to an agent without separating it is scary, with good reason.
How it is done today
What is known about each project is scattered across tickets, documents and people's memory. Finding a fact means asking whoever was there. And when that person is away, so is the fact.
How we solve it
- Each project's knowledge is gathered in a brain of its own, one per client.
- It is indexed every 15 minutes, so what is new is available without anyone loading it by hand.
- The agent working on a project queries only that client's brain.
- The isolation does not depend on the agent getting it right: it cannot see another project's notes, even by mistake.
Result
It is how we work every day: each project has its own brain and none of them mixes with another.
What it takes
Where what you know lives today
Access to each project's tickets and documents. What only lives in someone's head has to be written down first.
What goes in and what does not
Which documentation belongs in each brain. What should not be within an agent's reach stays out.
The boundaries between clients, in writing
Which projects belong to each client and where one ends and the next begins. The isolation is built on that list.
A person responsible for each brain
Whoever decides what goes in and who has access. Without that person, the brain fills up with things nobody has reviewed.
How it is built
Each project's knowledge is indexed in isolation, per client. Each client has its own brain, and the agent of a project only reaches its own.
Indexing repeats every 15 minutes. What is added or corrected in a project enters its brain on the next pass, with no one stepping in.
The isolation lives in the access, not in the agent's instructions. That is why it cannot see another project's notes even by mistake. Who can access each brain is decided by a person.
Frequently asked questions
Can an agent see another client's information?
No. Knowledge is isolated per client: the agent of one project cannot see another's notes, even by mistake.
How often is it updated?
Each project's knowledge is indexed every 15 minutes.
Who decides what goes into each brain?
A person. They also decide who can access it.
Related automations
Tickets up to date
The agent keeps ticket status up to date. A person sets priorities and decides.
Debate between models
Two AI models debate an idea and check every fact. A person decides with the verified facts in front of them.
Part of AI agent development