Open beta
Your company's context, in every teammate's AI.
Every conversation with your AI starts from zero. One Brain gives it the context nobody writes down: what the team decided, who said it and when. Your agent records it on its own while you work and brings it back when it matters.
Free during the beta · Sign in with your Google account
The app is in Spanish for now.
one brainwhy does the payments worker stop retrying after 2 attempts?
one brain · 1 team memory
That cap is deliberate, not a leftover. It was set after the duplicate-charge incident in March.
«Two retries on payments, hard stop. At five we double-charged 40 customers in one night. If a queue genuinely needs more, it comes through me.»
Priya Raman · Platform · Mar 12
Nobody writes documents
Your agent records what gets decided when a topic closes. No form to fill in, no wiki to keep alive.
Every answer has a name on it
Who decided it, and when. You stop reconstructing intent from commit messages six months later.
One topic, all of it, together
Every service, client or person is a tag. Ask about one and you get everything written about it, no matter which teammate wrote it.
It gets what you mean. And it won't hand you an old answer.
Ask in your own words and it finds what the team wrote in theirs. When two memories disagree, it flags them and someone on the team picks the one that stays.
AI with One Brain
"Starting today, freight to Mercado San Juan goes up to $38,000 per drop. Diesel went up."
AI with One Brain
"Any markdown over 5% needs Marcos Rivas to sign off. In July three orders went out at a loss."
AI with One Brain
"Order cutoff moves to 4 pm. Anything after that ships the day after tomorrow: the warehouse couldn't build the routes in time."
Getting in
Sign in with Google
Your brain exists immediately. No form, no card, no waiting to be approved.
Connect your AI
In Claude or ChatGPT you add it as a connector. In Claude Code or Codex, with one command. From then on every conversation starts knowing what the team is on.
Add your teammates
Everyone signs in with their own account and writes into the same brain. What one records, everyone has.
How it works
Four people on one team, each with their own AI, connected to One Brain: each AI sends what it writes down and gets back what the others wrote.
Mentions
From your AI you leave a to-do for a teammate. It shows up in their panel and in their AI when they start working. If your team linked a Telegram group, it also lands there with the context attached, and they can mark it done from the group.
OneBrain
@martin — Sofía Vega mentioned you
Almacén Los Andes signed the 12 months. Send them the new price list with the 15%.
Almacén Los Andes: price exception tied to a 12-month term
Questions
Does it work if I'm on my own?
Yes. You start alone and your AI stops forgetting. The day someone joins, everything you've built is already there for them.
Who can read what I store?
Your team: the people you invite. Each person also has a private memory only they can see. The Prophet team can also access it to give you support. No other company can read your brain. How we keep it
Can we run it on our own infrastructure?
Yes. Some companies are not allowed to keep internal knowledge on someone else's server, so the same product runs inside your own network: the app and the database are yours, and nothing crosses your perimeter. We set it up one company at a time, so write to us and we go through your requirements first. Talk to us
Does it read my source code?
No. One Brain stores what your team decides and learns — not your repository. You choose what gets recorded, and you can see and edit every entry.
Can I take my data and leave?
Yes. Whoever administers the brain downloads a file from Settings with the memories, meetings, entities, project sheets, mentions, syntheses, constitution and people. If you prefer, we can put it together for you. What it includes
Everyone writes into the same place, so the brain gets more useful every week: today's answers come out of what the team recorded over the last six months.