We launched on Product Hunt!We launched on Product Hunt! If you like Brain, please support us by following the launch.See the launch on Product Hunt (opens in a new tab)

Brain · The Brain Platform

An AI Knowledge Base for Your Internal Team References

Consult approved company documents through compatible AI clients, inspect source evidence and keep the original owners responsible for the guidance.

Fictional example · evidence you can inspect

What is the approved refund-review process for an annual plan?

The fictional support reference says an annual-plan refund request goes to the billing reviewer with the account context and original invoice reference. It describes a routing process, not an automatic refund entitlement. Open the source and involve the responsible reviewer before acting. This example contains no real customer or payment details.

Annual-plan support reference
Source excerpt · Annual-plan support reference

Route annual-plan refund requests to the billing reviewer with account context and the original invoice reference.

An AI knowledge base is useful when the maintained source is clear

An AI knowledge base organizes information so an AI tool can find context relevant to a question. Brain connects approved knowledge for compatible clients, keeping source evidence available for review.

An internal wiki can be well organized and still leave a teammate asking where the current answer lives. The wiki entry describes an older process, the approved document contains a later exception and a decision note explains why it changed. Brain is an internal knowledge workflow for that problem: consult approved team references, explain what supports the answer and let the reader return to the evidence.

Begin with documents the team already maintains rather than creating an article for every possible question. Connect an approved collection, preserve its owners and distinguish current guidance from an archive. A source-backed answer can locate the relevant passage across connected knowledge. It cannot resolve an unmarked conflict between two authoritative-looking policies without the responsible owner deciding which reference applies.

A maintained collection still needs care. Someone must update a changed policy, remove a superseded instruction or mark a historical decision clearly. Source processing and refresh behavior determine when a connected answer reflects that work. Test a known change in the pilot and inspect the resulting source rather than assuming an answer is current because the original file remains in a familiar shared folder.

The audience also matters. An internal team knowledge layer is different from a public customer help centre. Employees and scoped agents may need approved operating context, while customers need published support articles and an intentional delivery channel. Choose the collection and asker identities for the internal task, then test a restricted source before adding confidential information. Useful retrieval should not collapse every audience into one answerable archive.

Connect. Ask. Govern.

From scattered documents to a shared answer

  1. 01

    Connect the maintained operating references

    Choose the approved support guide, process pages and reference index. Name the owners and label historical material so the collection explains which document the team should use now.

  2. 02

    Ask the internal question with evidence

    Ask how to route a refund-review request or where a process step is recorded. Open the cited source and use the established owner for decisions or exceptions the reference does not settle.

  3. 03

    Govern the internal audience

    Review employee and agent scope separately from a customer help centre. Test a private source with synthetic material and repeat a question after a known policy update to check the connected behavior.

See the idea in action

ai knowledge base: questions with evidence

Fictional examples. These interactions do not query your Brain or test real permissions.

Try the example as

Open a question, then inspect its source excerpts.

Keep working in the AI tools you use

Brain is a knowledge layer reached through MCP, the Model Context Protocol. Claude, Cursor and Codex are examples of compatible clients in the existing product. Client support and setup differ, so check the current connection instructions rather than assuming every assistant has the same capabilities.

Connect a useful set of sources first

The current public setup describes Google Drive and Notion as connected sources, alongside documents you choose to add. Start with a focused collection and inspect the actual connection screen for availability. A tool appearing in a roadmap or illustration does not mean its connector is ready for your account.

Google Drive

Connect the documents your workspace needs.

Notion

Bring approved pages into a shared knowledge layer.

Your documents

Add a focused set of knowledge you own.

Keep access intentional

Keep internal references and private records deliberately separate

An internal knowledge base can serve several teams while respecting different source audiences. Decide which operating references belong in the ordinary collection and which cases require a narrower group. Review the actual connected identities and processing terms before adding sensitive material. Source evidence explains the process; it does not authorize a refund, employment action or other separate business decision.

Fictional content-blind access record
Actor
Example teammate
Action
Read approved source
Outcome
Allowed within the example workspace

Fictional metadata only. No document content is shown, and this is not a record from your account or proof of live enforcement.

Read the privacy policy

Supply the relevant operating context for the question

An intake-process question may need the current clause and its exception, not every page of the company wiki. Compare the source-backed answer with the approved full reference and measure actual context usage. Retain necessary definitions and include retrieval, model outputs and Brain charges when evaluating cost; a maintained source collection is valuable only when the answer remains suitable for the work.

Compare current plans and limits

Choose the knowledge system for its intended audience

Choosing an approach for this workflow
ApproachWhat to consider
An internal wikiA useful place to structure and maintain guidance. Owners still need to keep it current and make the accepted reference clear. Brain can consult approved wiki knowledge without claiming that every existing wiki should be replaced.
Customer help-centre softwareDesigned for publishing support information to customers and managing that audience. Evaluate it for public support delivery. An internal knowledge workflow needs different collections, identities and confidential-source boundaries.
Search the shared driveHelps locate files and inspect originals. It may leave the teammate responsible for comparing several passages or deciding which version applies. Clear names, owners and status remain useful whichever retrieval approach you choose.
Brain source-backed consultationLets compatible clients ask about approved connected references and inspect the supporting evidence. Keep the collection maintained and test source changes and access. It adds an internal question workflow rather than automatically publishing a customer portal.

ai knowledge base: frequently asked questions

What is an AI knowledge base?

An AI knowledge base organizes information an assistant can consult for a specific question. Brain applies that idea to approved connected documents and source-backed answers through compatible clients. This page concerns internal team references. Source owners, current guidance and intended access still need to be maintained; retrieval alone does not make every document authoritative.

How is it different from knowledge base software?

Knowledge base software can help publish and organize articles for employees or customers. An AI consultation layer helps a compatible client locate and explain relevant information from approved references. The approaches can work together. Choose based on your audience, maintenance responsibilities and access requirements rather than assuming every article system or AI tool serves the same purpose.

Do we need to migrate our wiki?

Start by connecting approved sources the team already maintains and testing a useful question. A complete rewrite into new articles may not be necessary, but connection can involve processing, indexing and storage. Review the source setup and privacy terms. Keep the original owners responsible for version status, accuracy and changes to the maintained references.

How does the knowledge base stay up to date?

The source owner updates the reference, and the connection’s processing and refresh behavior determine when that change becomes available to an answer. Test a known update and inspect the source in your pilot. Mark superseded guidance clearly. This page does not promise instantaneous synchronization or that retrieval can resolve every conflict between poorly maintained documents.

Can it answer from several tools at once?

Brain’s published source list includes Google Drive and Notion, so an approved collection can include references from both. Choose a process question supported by both references and check what each contributes. Verify readiness and access for the actual connections. A broader source list does not make every roadmap connector available or eliminate differences between source formats.

Who can see what in an internal AI knowledge base?

The intended answer depends on the asker’s identity, workspace and permitted source scope. Brain’s published model uses source-aware access for people and agents. Verify that behavior with an approved and restricted identity using synthetic documents, then repeat after an access change. An internal collection should not make private personnel or commercial records broadly answerable.

Can customers use the same knowledge base?

An internal team collection is not automatically a public customer help centre. Customer access needs an intentional product channel, approved published content and its own authorization decisions. This page does not promise a customer portal or public chatbot. Start with the internal task and keep private operating knowledge separate from information the organization has approved for external use.

Make one internal process easier to understand

Connect its approved references, ask a question your team can verify and inspect the evidence. Keep the original owner in charge of updates as the collection grows.