Microsoft Teams Integration

Brainfish in your Teams channels. Sourced answers, no new tab.

@mention Brainfish in any Microsoft Teams channel and get a grounded, cited answer right there in the thread — without leaving the conversation. It’s the same pattern teams using Slack have used for months, now available for enterprise Teams environments.

In addition to answering questions, Brainfish can also turn useful Teams threads into draft knowledge base articles for human review and publishing.

What You Can Do

• Get answers in any channel — @mention Brainfish and receive a sourced, cited answer in-thread within seconds
• Follow up in the thread — ask follow-up questions without re-mentioning Brainfish each time
• Surface knowledge where your team already works — no new platform, no tab-switching, no behaviour change required
• Rate answers — Helpful / Not Helpful feedback per answer, feeding directly back to your analytics
• Generate knowledge from Teams threads — @mention Brainfish in a thread to generate a draft knowledge base article from that conversation (same workflow as Slack)

How It Works

Brainfish can be added to Microsoft Teams channels where conversations happen (e.g., support, product, internal help).

When someone @mentions Brainfish, it can:

1. Read the relevant conversation context (including the full thread where it was mentioned)
2. Search your connected knowledge base and generate a grounded, cited answer
3. Post the answer in-thread, visible to all channel members
4. (Optional) Generate a draft knowledge base article from the thread for human review

Recommended Workflow (Internal vs Public Knowledge)

Many teams use Teams to capture internal know-how (support playbooks, engineering context, onboarding answers) while keeping the public Help Centre customer-safe and curated.

Ongoing capture from Teams (curated, @mention-based)

1. In Teams, identify a valuable thread (Q&A, troubleshooting steps, decisions, how-to guidance).
2. Reply in the thread so Brainfish captures the right context.
3. Tag @Brainfish and request article creation, for example:
   • @Brainfish create an article from this thread
   • @Brainfish create knowledge from this
4. Find the draft in Brainfish — open Knowledge → Articles → Knowledge waiting for your review.
5. Review, edit, and publish — drafts are never auto-published.

Important: The Teams integration is @mention-based and does not passively monitor channels. You control what is captured.

Where to store the article (Internal vs Public collections)

After you review a draft, place it in the right collection:

• Internal collections
  • Visible only to your team
  • Not surfaced in the public Help Centre or widgets
  • Can still inform AI-generated answers across your Brainfish agents
• Public collections
  • Customer-facing
  • Visible in your Help Centre and widgets

Set a collection’s visibility (Internal vs Public)

1. Go to Knowledge → Managed CMS (or the Articles/Collections view in your workspace).
2. Select the collection.
3. At the top, click the visibility/permissions button (shows Internal or Public).
4. Choose the appropriate setting (updates immediately).

Generate Knowledge from Teams Threads

Useful Teams threads — troubleshooting steps, product Q&A, onboarding answers — can be turned into draft knowledge base articles without leaving the conversation.

When to Use This

Use Teams article generation when:

• A Teams thread contains a helpful Q&A, troubleshooting walkthrough, or explanation worth preserving
• Your team answers a question in Teams that is likely to be asked again
• Product, support, or internal-help channels contain repeatable guidance you want searchable in your knowledge base

How to Create an Article from a Teams Thread

1. Reply in the thread so Brainfish captures the right context.
2. Tag @Brainfish and request article creation, for example:
   • @Brainfish create an article from this thread
   • @Brainfish create knowledge from this
3. Wait for processing — Brainfish analyses the entire thread and generates a draft. This typically takes 30 seconds to a few minutes; you can navigate away while it runs.
4. Find the draft in Brainfish — open Knowledge → Articles → Knowledge waiting for your review.
5. Review, edit, and publish — drafts are never auto-published. Review for accuracy, edit as needed, and click Accept to publish.

Once accepted, the article is searchable in your knowledge base and immediately available to your AI agents.

Key Features

• Full thread context — Brainfish reads the complete thread (not just the message it’s tagged in), so drafts capture the actual answer and reasoning
• Smart deduplication — when similar content already exists, Brainfish suggests an update to the existing article instead of creating a duplicate
• Same workflow across Slack and Teams — one review queue and one editorial process, regardless of where the thread originated
• Included in standard pricing — no add-on required

Historical Teams Conversations (Backfilling Older Threads)

The Teams integration processes content going forward (via @mentions). There is no native one-click retroactive import of historical Teams threads via the integration.

If you want older Teams conversations (from before setup) to inform Brainfish answers, choose one (or combine multiple) approaches below.

Option 1: Bulk import via Catalog API (fully automated, requires developer work)

Use this when you want broad historical coverage without manual review.

High-level flow:

1. Export historical Teams conversations using the Microsoft Graph API.
2. Create a Brainfish catalog for the source.
3. Push threads as formatted files into Brainfish via the Catalog update API.

Outcome:

• Historical content becomes searchable context that can inform Brainfish answers.
• Best suited for large volumes of content.

Option 2: Selective curation using @mentions (manual, no dev work)

Use this when you want only the highest-value historical threads.

Process:

• Have team members revisit key threads and run @Brainfish create an article from this thread.
• Review and publish into an Internal collection.

Option 3: Brainfish MCP (semi-automated curation, no dev work)

Use this when you want curated, structured articles without building an API sync.

How it works:

1. Export or copy historical Teams threads.
2. Paste them into Claude with the Brainfish MCP connected.
3. Ask Claude to draft and create articles in Brainfish (typically targeting an Internal collection).

Important limitation:

• Brainfish MCP is not a Teams connector; someone must provide the historical threads to Claude.

Choosing the Right Backfill Approach

• Catalog API: Best for comprehensive, automated historical ingestion (requires dev effort)
• Brainfish MCP: Best for efficient, semi-automated article creation with human-curated inputs (no dev work)
• @Brainfish in Teams: Best for ongoing capture, thread-by-thread, as new conversations happen

Requirements

• Microsoft Teams (Business or Enterprise)
• Brainfish app installed in your Teams workspace by an admin
• Brainfish added to the relevant channels by a Teams admin
• Channels must be configured in Brainfish admin settings

Setup

Teams integration is configured at the account level. To get started:

1. Contact your Brainfish account team or email support@brainfi.sh
2. Your account team will assist with installing the Brainfish app in your Teams workspace
3. A Teams admin adds Brainfish to the relevant channels
4. Optionally, configure which channels Brainfish responds in from Brainfish Admin Settings

Once configured, your team can start @mentioning Brainfish immediately — no training or onboarding required.

Analytics

Teams activity is tracked separately in your Brainfish analytics dashboard, so you can see which channels are most active and what questions are being asked.

Privacy and Permissions

Brainfish does not passively monitor Teams. It only processes conversations when it is explicitly tagged with @Brainfish.

• For answers: Brainfish processes the question and the relevant in-thread context.
• For article generation: Brainfish processes the full thread where it was tagged to draft an article for your review.

Your Teams admin must approve the Brainfish app before it can be installed. For compliance documentation, contact support@brainfi.sh.

Frequently Asked Questions

Is this the same as the Slack integration? Yes — it follows the same proven @mention-to-answer pattern and the same @mention-to-generate-knowledge pattern. Same admin governance model, same knowledge base, same cited answers, and the same review queue.

Can I control which channels Brainfish responds in? Yes. Channel scoping is admin-controlled from Brainfish settings — Brainfish only responds in channels you configure.

Can Brainfish create articles from direct messages? Article generation works best in channels where Brainfish has been added. DM-based generation isn’t supported.

What if the thread contains sensitive information? Drafts require manual review before publishing. You can edit or discard any suggestion before it goes live — nothing reaches your knowledge base automatically.

Will Brainfish create duplicate articles? No — Brainfish is designed to suggest updates to existing content where relevant, rather than always creating a new article.

What if Brainfish can't find a confident answer? Brainfish will say so rather than guessing.

Does this work for internal team questions as well as customer-facing support? Yes — Teams integration is ideal for internal knowledge sharing: product questions, onboarding, or support team lookups mid-ticket.

References

• Microsoft Teams Integration: /articles/microsoft-teams-integration-GeicuiQvw9
• Collections 101: /articles/collections-101-FTTfDEYDaH
• Bulk Upload Knowledge into Brainfish: /articles/bulk-upload-knowledge-into-brainfish-4s8T85GM7w
• Brainfish MCP — Use Brainfish Inside Claude: /articles/brainfish-mcp-use-brainfish-inside-claude-sykvRMcDrV
• (Related pattern) Sync Slack Channel Messages into Brainfish: /articles/sync-slack-channel-messages-into-brainfish-39RYXgrrg1