Shared.
Everyone in the group reads the same thing. So does every AI agent working for them. No more private versions.
A little more common ground.
Shared knowledge your team can fall back on.
Treo is your organization's shared knowledge, kept current by AI and open to every person and AI agent you work with. Every answer shows where it came from.
The one brain your whole team shares.
Everyone on the same page.
TreoIllustrative example
What discount can I offer on a three-year contract?
Your team’s shared answer
Up to 20%.
For contracts of two years or more.
Shorter contracts stay at 15%.
Good work starts with shared understanding.
The familiar problem
What does your team fall back on today?
Most teams have no good answer. What is written down was true on the day it was written. What is true today lives in people’s heads, inboxes and private AI chats.
Two people ask the same question and get two different answers.
Each person holds their own version. Now each person's AI agent does too.
The document everyone points to was last right six months ago.
Keeping it current is nobody's job, so it goes out of date while still looking finished.
The person who knew has left, or is on holiday.
The decision is on record. The reason for it is not.
So people ask whoever seems to know, and hope.
Meet Treo
One body of knowledge.
One body of knowledge.
Shared, current and checkable.
Current.
When you add a new document, your AI agent works it into what is already known. Pages are updated as part of normal work, not as a chore someone has to remember.
Checkable.
Every claim links to the source behind it. The original is kept exactly as it was. When two sources disagree, you see both.
You get your first brain at birth. A second brain is your own notes. Treo is the one your whole team shares.
Follow the evidence
Follow one question
Follow one question
from answer to source.
A leadership team keeps its pricing knowledge in Treo. Here is what happens over one week.
Step 1 of 5
Something new comes in.
The sales lead asks their AI agent to add the notes from the leadership meeting. The notes are saved as they are and become searchable.
New source · original preserved
Leadership meeting notes
7 October · Added by the sales lead
20% was agreed for contracts of two years or more.
Illustrative example · Select each step, then open a source.Sign up for the beta
Trust, with a trail
Built so that you can check,
Built so that you can check,
not just trust.
Originals are never overwritten.
A new version of a document is saved next to the old one, not on top of it.
Every change has a name and a reason.
You can see who changed a page, which agent they used, and why.
Nothing is lost.
Every earlier version of a page stays readable.
Disagreements stay open until someone decides.
An agent can fix a typo on its own. It cannot quietly rewrite a conclusion your team has agreed on.
Each space has one audience.
Everyone in a space can read everything in it, including the sources. There are no hidden corners, so there are no surprises.
A common ground
Any group that needs to know
Any group that needs to know
the same things,
and keep knowing them.
What you fall back on today
Small and medium-sized companies
The founder's memory, and the one person who has been there longest.
An everyday example
The office manager who ran supplier contracts for eight years is leaving. During the last month they add the contracts, the renewal dates and their own notes on each supplier. A new colleague later asks: "When do we need to give notice on the cleaning contract, and is there anything I should know?" The answer gives the date, links to the contract, and includes the note that this supplier has always agreed to a lower price when asked.
What you can rely on
Know-how that stays when people move on, and new staff who can find answers on day one.
Illustrative scenarios, not customer stories. Space availability and outside-member access will be confirmed at launch.
Fits the way you work
Three steps, using the
Three steps, using the
AI tools you already
have.
Connect.
Connect Treo to your AI agent. Sign in and choose what the agent may do. Use any agent you wish; plugins for ChatGPT and Claude will be supplied.
Add what matters.
Ask your agent to add a document, meeting notes or a decision. The original is saved, and your agent works it into the shared pages.
Ask.
Anyone in the space can ask their own agent and get the same answer, with sources. Or open the browser to read, search and follow any claim back to where it came from.
There is no new chat window to learn and no editor to master. You keep working the way you already do.
For the technically curious: read how it is built
Treo builds on the idea of a wiki that an AI model writes and maintains from your raw sources. It is designed for many people at once, with permissions, version checks and a record of every change. Treo stores, checks and searches; the AI work is done by the agent you already use.Read how it is built.
Clear boundariesStraight answers
Straight answers
about the limits.
It does not listen in.
Nothing from your private chats is added unless you choose to add it.
It does not update itself in the background.
Knowledge is kept current by your team's agents as you work. If nobody uses it, nothing changes.
It does not decide for you.
When sources conflict on something that matters, a person with the right to decide makes the call.
It does not promise the AI is always right.
It promises that you can check. Every claim links to its source.
Questions, answered
A little more clarity.
General
How is this different from a wiki or a shared drive?
A wiki depends on people remembering to update it. A shared drive stores files but does not connect them. With Treo, your AI agent does the upkeep as part of normal work, and every page links back to the documents behind it.
How is this different from the memory in my AI agent?
That memory belongs to you alone. Your colleague's agent knows different things. Treo is one memory that everyone's agent reads from and adds to.
How does it stay current?
Each time someone adds a source, their agent updates the pages it affects. Agents can also review a part of the knowledge and fix small problems. If your AI tool supports scheduled tasks, you can run that review on a timetable.
Which AI agents can we use?
Any agent you wish. Plugins for ChatGPT and Claude will be supplied.
What can we add?
Documents, meeting notes, decisions, policies and anything else your agent can read as text. If it cannot read a file, it tells you and does not guess.
Can the AI get things wrong?
Yes. That is why every claim carries a link to its source, every change is recorded, and every earlier version can be read again.
Who can see our knowledge?
Only the people you add to a space. Each organization's data is kept apart from every other organization's.
Does Treo run AI on our data?
No. Treo stores, checks and searches. The AI work is done by the agent you already use, under the terms you already have with that provider.
Who can change things?
There are four levels: read, contribute, decide and administer. Readers can ask and browse. Contributors can add sources and update pages. Deciders settle open questions. Administrators manage who has access.
Can we edit pages in the browser?
No. The browser is for reading, searching and checking history. Changes go through your agent, so every one of them is checked and recorded.
How long does it take to set up?
About five minutes. You connect your AI tool, sign in, and add your first document. The knowledge grows from there.
Can we get our knowledge out again?
Yes. You can export the whole knowledge base as Markdown files at any time.
What does it cost?
The price will be set later. It will probably be around €9 per user per month.
For small and medium-sized companies
We have no time to write things down. Will this help?
Yes. You do not write pages. You hand your agent what you already have, such as a contract, an email or notes from a call, and it does the writing.
We are only ten people. Is it too much for us?
No. It is built for small groups. A small team often has the most to lose when one person leaves.
For leadership teams
Can we keep leadership matters apart from the rest of the company?
Each space has one audience. The number of spaces per account will be confirmed at launch.
What if we disagree about what was decided?
Then the disagreement is recorded as an open question, with the evidence on each side, until someone with the right to decide settles it.
For boards
Can directors check what management tells them?
Yes. Every summary links to the original paper or minutes, so a director can read the source and not only the summary.
What happens when a director leaves?
Their access ends. The knowledge, the decisions and the reasons stay for those who follow.
For shared projects
Can people from different organizations use the same space?
Shared project spaces are part of the product direction. Details on inviting outside members will be confirmed at launch.
What happens to the knowledge when the project ends?
The knowledge stays readable for as long as you keep the space. You can export it as Markdown files at any time.
For client-facing firms
Can one client's knowledge leak into another's?
Each space has one audience, and knowledge is kept within that space. The number of spaces available per account will be confirmed at launch.
Can the client have access too?
Everyone added to a space can see everything in it, so decide what belongs there first. Details on inviting outside members will be confirmed at launch.
Give your team something
to fall back on.
One shared body of knowledge, kept current as you work, where every answer shows its source.
Sign up for the beta