Questions, answered.
What Team0 is, how a pilot works, who sees what, and how it runs in your own cloud.
The basics
What is Team0?
Team0 is a brain for your company’s AI agents. It learns from your people’s work by itself, keeps track of what changed, who promised what and what is still open, and gives that to every agent your people use. There is a brain for each person and one for the company. It runs in your own cloud, or we host it for you on a server of its own.
Who is it for?
Companies whose work passes between people: clients, projects and decisions. It fits best where people already use AI agents and lose time waiting on what a colleague knows.
What problem does it solve?
Every person learns something new each day, and their agents only know what that one person told them. A client changes a requirement in one meeting, and a colleague’s agent keeps working from the old version. Team0 lets each person’s agents know what their colleagues know, with the source attached.
Do I have to write anything down or keep it organised?
No. Team0 learns from the work people already do, like meetings, email, work tools and the conversations their agents save back. There are no notes to write and nothing to tidy.
Isn’t this just search over our documents?
Search finds what was written and leaves the agent to guess which version holds. Team0 learns from the work itself, including meetings and email, and keeps track of what changed. Ask for a price and search returns three emails; Team0 returns the current price, the change coming next month and the proposal nobody adopted, each with its source.
Isn’t this the same as the memory in ChatGPT or Claude?
That memory belongs to one person in one product. It only knows what that person told it, and switching agents starts over. Team0 learns from everyone’s work, with each person deciding what is shared, and every agent reads the same brain.
My agents already share a notes file. Why do I need this?
It works at first. Then the file grows, key details get summarised away, and nobody knows which version still holds. Someone ends up maintaining it, and every agent reads a slightly different version of the company. Team0 keeps itself current from the work.
Why “understanding” and not “memory”?
Memory stores what was said. Understanding keeps track of what is still true. When a deal changes hands, the new owner replaces the old one and the old one moves to history. A customer asking for something is kept as a request, not a promise. Things nobody knows yet stay marked as unknown instead of being guessed.
Pilots and setup
How do we start?
Book a call. We pick one job where waiting on a colleague costs you time, install Team0 around it, and you judge it on ten of your own questions.
What don’t you do?
We don’t build custom agents or workflows as part of Team0, and we don’t replace Claude or ChatGPT. Team0 sits next to them, as one brain for all your agents, whichever company made them.
Who does the setup?
We do. We install Team0 in your AWS or Google Cloud account, or run it for you on a server of its own, then connect your people, their sources and their agents. After that, it builds and maintains itself.
How do people get access?
By invite only. An admin invites each person by email and can pause anyone at once.
Which sources can I connect?
Each person connects Gmail and Google Calendar, and meeting transcripts can be added. Work tools like HubSpot, Notion and the other tools your people use connect too. The company’s brain has its own sources too: a shared inbox and calendar, Drive folders, the CRM and work tools, company documents and the Slack channels an admin picks.
Does it learn from Slack?
From the public channels an admin picks, for the company brain. It reads them like meetings, keeps work facts and skips chatter. Message text is kept 7 days, then only the line behind each fact and a link. Each person can opt out, and the app posts a notice in every channel it joins. It never reads private channels or direct messages.
Can it learn from my meetings?
Yes. Meeting transcripts are added to the brain of the person who was there. What is said in a meeting is kept as “heard” until something confirms it.
How soon is it useful?
Each person’s brain starts filling in as soon as they connect, from their recent email, calendar and meetings. After that, email and calendar update hourly, agent turns arrive live, meetings arrive when they end, and other tools about daily. The test on your own questions shows what it adds on your own questions.
Your agents
Which agents work with Team0?
ChatGPT, Claude, Claude Cowork, Claude Code, Codex, OpenClaw, Hermes, Town, Lindy, monday.com and Base44. Most agents that support MCP can connect too.
How do I connect an agent?
We connect them during setup, and each person can add more later from Team0. Each agent gets its own access, which can be changed at any time.
Do my agents use Team0 automatically?
Yes. Claude Code, Codex, OpenClaw and Hermes read and save on every turn, through a plugin. The others, like ChatGPT, Claude, Lindy and monday.com, connect over MCP with standing instructions that tell them when to read Team0.
Do my agents save what I tell them?
By default, yes. After each turn, Team0 receives your message and the agent’s final reply. It doesn’t open your files or watch the agent’s tools, but anything inside a message comes with it, including what you paste. Team0 pulls out what matters and deletes the conversation itself after 7 days. What it learned stays, with the line each piece came from. What an agent says is kept as “the agent said”, never treated as fact on its own.
Can I stop an agent from saving, or cut it off?
Yes, per agent. One switch stops an agent from saving while it keeps reading. Another cuts it off completely, and its next request is refused. Other agents are not affected.
Can I see what an agent read?
Yes. Every read is on the record: the exact request and what Team0 returned. The full result is kept for 24 hours, the read’s content for 24 hours to 30 days (each person sets it), and the record of the read for 30 days.
Can my agents talk to each other?
Yes. Ask one of your agents to message another of your agents, or a coworker’s agents when that coworker allows it, and the message waits in that agent’s Team0 inbox until it next checks. Replies come back to the agent that asked. It can be switched off per agent.
Can my agent find out what a colleague knows?
Yes, with that colleague’s permission. Your agent can check what they know about a client or company, and the answer names whose it is. Their private work is never part of it. Nothing is kept, and they see that you asked.
Is there a limit on how many agents I connect?
No. Each person can connect every compatible agent they use.
Can I build my own agent or product on Team0?
Yes. Any MCP client can connect to Team0’s MCP server, and products can use the HTTP API.
How the brain works
What does Team0 keep track of?
The people and companies in your work, commitments, decisions, open questions and what changed. Every statement keeps who said it, when, and the source it came from.
What happens when something changes?
The new version becomes current and the old one moves to history instead of sitting next to it. A correction made once reaches every agent you connect.
What if it gets something wrong?
Every statement shows where it came from, so it can be checked. Correct it or remove it once, and every agent reads the fix. Extraction can misread, which is exactly why every statement shows its source.
Does it trust everything it reads?
No. A request is not a commitment, something heard in a meeting stays “heard” until confirmed, and nothing an agent or a model says is trusted automatically. Silence never counts as something being done. What an agent reports it did is kept as reported, not confirmed.
Does it know everything from day one?
No. It knows what connected sources and agents have shown it, and it gets fuller as they sync. When it doesn’t know something, it says so instead of guessing.
What is the company’s brain?
A full brain of its own: its own sources (shared inbox and calendar, Drive folders, CRM and work tools, documents, Slack channels an admin picks), the work facts people choose to share, and the same truth checks. Not search: it knows what changed, who promised what and what is still open. Company knowledge flows down ahead of time into every brain that knows the client, read-only and labelled “from the company brain”. Each shared fact names who shared it, so you can see who knows what.
Who owns the company’s brain?
The company does. Whoever runs Team0 for you decides who may change it. Each person owns their own brain and chooses what they share with the company.
Privacy and security
Do you train AI models on my data?
No. We never train AI on your data, and your AI accounts run under their providers’ business terms.
Who sees what?
Each person’s brain is their own. Nobody else reads it: not the admin, not coworkers. Sharing is off until each person turns it on. Only work facts are shared, and from email only company-domain mail, each with the sharer’s name, and the quotes and evidence stay in their brain. They can withdraw a shared fact at any time, and it is gone within a minute. Each agent sees only what it is allowed to.
Can Team0 see our data?
Only through the operator access you grant us to install and maintain it, and you can revoke that access at any time. We don’t browse customer content.
Which AI models process my data?
Team0 calls AI models with your company’s own keys, under your own terms with those providers.
Do you have SOC 2?
No. When it runs in your cloud, your data never sits on our servers. When we host it, it gets a server of its own. We go through the security details on a call.
Where is my data stored?
In your own AWS or Google Cloud account, in the region it is installed in. If we run it for you, it sits on a server of its own.
What can each agent see?
Only what it is allowed. Its access is applied inside the query that reads the brain, so what it may not see is never retrieved. Reading the company brain is on when you connect an agent, and you can switch it off for any agent.
What happens if I disconnect a source?
Team0 stops reading it. What it learned before stays until you remove it.
Can I delete my data?
Yes. Remove any statement and it stops being used for new reads within a minute. When Team0 runs in your own cloud account, the data stays under your control. When we host it, it sits on a server of its own.
What happens when someone leaves?
The work facts they shared stay with the company. Their private brain is erased, with no waiting period.
Pricing and accounts
How much does it cost?
It depends on the size of the team and the job. We go through it on the first call.
I already have a Team0 account. What changes?
The earlier Team0 app has been retired. Reach us through the support page and a person will help you.
How do I get help?
The support page has every way to reach us, and a person answers.