Public source · Dataset
Published, versioned capabilities
The registry exposes integrations, tools, requirements and limits extracted from Meteor’s public catalogue, with a version and verifiable fingerprint.
Operational Company Brain
Meteor connects conversations, contacts, collections, documents and MCP data so employees and agents can query and act with context, permissions and traceability, including on WhatsApp.
Direct answer
An internal AI chatbot is the interface employees and teams use to query the knowledge they are allowed to access. In Meteor, that interface reaches the Company Brain: the company’s shared operational context. It connects collections, CRM, conversations, processes and business systems through MCP. Mets query that information, run authorised tools, update context and leave a trace so the next person, agent or automation can continue the work. Authenticated members are governed through roles and scopes; on WhatsApp, each Met operates only on the collections and permissions configured for its channel.
This is not a future promise: it is the architecture Mets already use. You can see operational use cases, read the business and technical guide, inspect the controls operating today and explore AI process automation.
Verifiable evidence
Each source proves something different. The registry and technical article are public; the Mega Alfa Trans case preserves its scope and states that its public video is still pending.
Public source · Dataset
The registry exposes integrations, tools, requirements and limits extracted from Meteor’s public catalogue, with a version and verifiable fingerprint.
Public source · Controls
The guide documents collections, permissions, live data, context feedback, approvals and traces that already form part of the operation.
Inspect operational controlsCustomer experience · Public video pending
Its team describes notifications, document intake and maintenance context on WhatsApp. The case separates that testimony from technical capability and publishes its limits.
Read scope and source statusArchitecture
Value appears when sources, governance and the channel work as one journey.
Collections, contacts, conversations, procedures and systems connected through MCP. Operational facts are queried at their system of record.
Members use roles and scopes. Every channel Met receives only its configured collections and fields, with read or write permission.
People ask in the channel they already use. The Met queries connected tools, updates authorised context and leaves an execution trace.
Role-based governance
Roles govern workspace members; channel permissions determine which company knowledge the Met may use on WhatsApp.
May access: Catalog, authorised commercial terms, CRM, quotes and opportunity status.
Stays out: Payroll, internal evaluations and restricted financial information.
May access: Procedures, policies, permitted contact history and the relevant case or order status.
Stays out: Margins, the full receivables book and unrelated customer information.
May access: Inventory, orders, procedures, owners and exceptions in the connected operation.
Stays out: Commercial or personal information that is not required for the process.
May access: The metrics and summaries its permissions allow, based on the current state returned by each connected system.
Stays out: Anything outside the workspace or outside the scopes approved for the account.
Comparison
The labels sound similar. The difference is who gets access, which source wins and whether the system can move the work forward.
| Category | What it organises | Who accesses it | Primary outcome |
|---|---|---|---|
| Personal second brain | One person’s notes and memory | The person who built it | Remember and connect information |
| Knowledge base | Documents, policies and FAQs | Teams with repository access | Find published content |
| Internal chatbot | Loaded content and selected integrations | Anyone who can open the chat | Hold a conversation and guide |
| Company Brain | Live context, collections and connected operational data | Members and channels within their configured scope | Answer, act and feed context back into the operation |
Available foundation
Company configuration
Published limits
These limits explain how Meteor works and keep connection, access and memory from being mistaken for unlimited permissions.
Connecting an MCP does not automatically copy its complete database into Meteor. Mets query live data, while agreed processes may save results to contacts or collections.
A WhatsApp number does not automatically inherit an authenticated member role. The channel controls the Met through collection-level read and write permissions.
Workspace RBAC governs users; the channel allowlist governs the collections an agent may use on WhatsApp. These controls complement each other.
A document does not replace a live ERP, CRM or store query when the fact changes. Meteor queries the system that owns the current state.
Meteor does not depend on presenting everything as RAG. Its operational core combines conversational context, structured data and connected tools; errors and exceptions still need controls.
Creating, updating, paying, publishing or modifying records requires an authorised tool and, when risk demands it, confirmation or human approval.
Frequently asked questions
In Meteor, the internal chatbot is a conversational entry point to the Company Brain. Each Met receives only the collections, fields and tools authorised for its channel; it queries documents and live CRM, ERP or connected-system data, answers with that context and records its actions. A WhatsApp number does not inherit an authenticated member role by itself.
A Company Brain is a business knowledge layer that connects sources, permissions and tools to answer questions with company context. Unlike a document folder, it must know which source to query, who may see it, whether it is current and which action may follow.
They share the idea of preserving and retrieving context. A second brain is usually personal; a Company Brain must handle identity, permissions, shared sources, ownership, auditability and boundaries between teams. That governance makes it usable inside a company.
Yes. A Company Brain can operate on WhatsApp through a Met configured for that channel. The operator selects which collections it may read or modify. Authenticated member roles and scopes are managed separately; an incoming number does not receive internal privileges automatically.
Yes, within the scope of available integrations. Collections preserve structured knowledge and company context; CRM, ERP, stores and scheduling systems are queried through MCP when an answer depends on changing state. Agents and automations may update the agreed context.
For authenticated members, Meteor applies workspace roles and scopes. On a channel such as WhatsApp, the operator configures which collections the Met may read or modify; an incoming number does not inherit a member role. Tools receive that scope and their execution is recorded.
Choose one process, organise its data into collections, connect the required systems and configure which context, fields and tools each Met receives. Then test real questions and actions, inspect traces and expand scope without migrating the whole company at once.
Start with the highest-value process, connect its sources and define each Met’s scope. Then expand questions, actions and teams on the same operational architecture.