Build your company’s brain.
You re-explain your company in every AI chat — and the day you forget, the answers are confidently wrong. Write it down once, in one place Claude, ChatGPT and Gemini all read.
Claude
ChatGPT
Gemini
Paperclip
OpenClaw
HermesWritten for your team. Structured for your agents.
Your knowledge shouldn’t need a translation layer. A Tatara document reads like a document to a person, and comes back as clean structured data to an agent. Same file. No export, no sync, no second copy.
Ideal customer profile
Who we sell to, who we do not, and the signals that tell them apart.
- sales
- sales/positioning
- acquisition
Who this is for
Operations and knowledge leads at companies of 10–50 people who have already started running AI agents against their own material. Pricing conversations follow Pricing playbook.
Disqualifiers
- Fewer than three people who write anything down.
- No agent in production — nothing to ground yet.
{ "document": { "id": "d0c8f3a1-5e42-4b77-9a10-6f2b8c4e1d93", "path": "Sales/ICP.md", "title": "Ideal customer profile", "content": "## Who this is for\n\nOperations and knowledge leads at companies of 10–50 people who have\nalready started running AI agents against their own material. Pricing\nconversations follow [[Pricing playbook]].\n\n## Disqualifiers\n\n- Fewer than three people who write anything down.\n- No agent in production — nothing to ground yet.", "type": "reference", "tags": [ "sales", "sales/positioning", "acquisition" ], "metadata": null, "links": [ { "title": "Pricing playbook", "path": "Sales/Pricing playbook.md", "resolved": true } ] } }
Your agents connect in about a minute.
Tatara speaks MCP — the open protocol Claude, ChatGPT and Gemini already use. Add it once; every one of them reads the same brain.
- 01
Add Tatara in your tool
Claude, ChatGPT or Gemini — one entry, once.
- 02
Approve access
You sign in and grant the connection. Nothing else to configure.
- 03
The agent reads
It searches your brain and gets the document back, structured.
{
"query": "who do we sell to"
}{
"query": "who do we sell to",
"results": [
{
"id": "d0c8f3a1-5e42-4b77-9a10-6f2b8c4e1d93",
"path": "Sales/ICP.md",
"title": "Ideal customer profile",
"snippet": "Operations and knowledge leads at companies of <b>10–50</b> people…",
"type": "reference",
"tags": [
"sales",
"sales/positioning",
"acquisition"
]
}
]
}A document when you want to read. A board when you want to plan.
Every document is plain Markdown underneath. Switch how you work in it — a filterable base, a typed table with formulas, a kanban board, a task tracker — without converting anything or leaving the file.
| Title | Type | Tags | Updated |
|---|---|---|---|
| Ideal customer profile | reference | sales, acquisition | 2026-07-21 |
| Pricing playbook | playbook | sales, pricing | 2026-07-18 |
| Discovery call script | process | sales, discovery | 2026-07-11 |
Your knowledge, in the shape the industry is standardising on.
Underneath, a brain is plain Markdown with YAML frontmatter — one file per concept, links forming a graph. It’s the same open shape Google’s Open Knowledge Format now formalises as an open standard: no proprietary schema, nothing you couldn’t open in any editor.
What a format can’t carry is trust. Every agent write into a brain is governed — type, title, description and tags are required and steered to your taxonomy — and every save is stamped with who, or what, made it. An agent provably cannot pass itself off as a person.
It stops being a folder the moment there’s a second person.
Roles and permissions. Invitations. People who write and people who only read — unlimited viewers on every plan, so the whole company can use the brain while a handful of people keep it true.
Flat, per-brain pricing.
Unlimited agent access on every plan — we never meter reads. Pick the seats and document headroom your team needs.
Founding pricing — 30% off your first 12 months — until 30 September 2026.
Starter
$49/mo
- 5 author seats · unlimited viewers
- 10,000 documents
- Unlimited agent access
Team
$149/mo
- 10 author seats · unlimited viewers
- 25,000 documents
- Unlimited agent access
Solo and Enterprise, annual billing, and the full comparison live on the all plans page.
Build yours.
I’m Angus. I built Tatara because I was pasting the same company context into every AI chat, and watching good agents produce confidently wrong work whenever I forgot.
It’s early, and I’d rather tell you where the edges are than sell around them. The workspace, the MCP connection and the render formats are live. The maintenance loop — the part that keeps a brain current on its own — is designed and not built.
If you’re putting a company brain in place, email me and I’ll help you set it up.
