🔓 Unlock all 10,000+ workflows & prompts free Join Newsletter →
✅ Full access unlocked — explore all 10,000 AI workflow and prompt templates Browse Templates →
Home › n8n › Workflow
January 21, 2026

Teams to Mem.ai, meeting notes you can actually use

Teams to Mem.ai, meeting notes you can actually use - n8n workflow automation template
Lisa Granqvist Partner Workflow Automation Expert
Teams to Mem.ai, meeting notes you can actually use - n8n workflow automation template
Get a free AI assessment → Get a free AI assessment → ⬇️ Use template

You finish a Microsoft Teams meeting, and then the real work starts. Someone has to find the transcript, skim for decisions, turn loose talk into action items, and chase people for follow-ups.

This Teams Mem.ai automation hits project leads first. But marketing managers and ops folks feel it too, especially when meetings stack up and “I’ll send notes” quietly never happens.

This workflow turns Teams transcripts into clean summaries inside Mem.ai, then drafts follow-up emails in Outlook so you can keep momentum without living in copy-paste hell.

How This Automation Works

The full n8n workflow, from trigger to final output:

n8n Workflow Template: Teams to Mem.ai, meeting notes you can actually use

The Problem: Teams Meeting Notes Don’t Survive the Week

Teams gives you the raw ingredients (recordings, chat messages, transcripts). It does not give you the usable version. So notes end up scattered across places that don’t talk to each other: a Teams chat thread, a random Word doc, somebody’s inbox, and maybe a half-finished task list. Then a week later, you’re trying to remember why a decision got made, who owned the next step, and what the actual deadline was. The waste isn’t just time. It’s the rework and second-guessing that comes from “meeting memory” being unreliable.

It adds up fast. Here’s where it breaks down in real life.

  • Someone has to manually pull the transcript from Microsoft 365, and that alone can take about 10 minutes per meeting.
  • Action items get paraphrased differently by different people, which means the team argues about what was “actually agreed.”
  • Follow-up emails are delayed because writing them feels like starting from scratch, especially after a long day of calls.
  • When notes aren’t stored in a searchable knowledge base, decisions get repeated in new meetings and the calendar fills with avoidable syncs.

The Solution: Turn Teams Transcripts into Mem.ai Summaries + Outlook Drafts

This n8n workflow watches for new Microsoft Teams meeting data (recordings and transcripts) within a chosen time window, then pulls the transcript details through Microsoft Graph. From there, it cleans and structures the content so an AI model can do what it’s good at: summarizing long, messy conversations into something you’ll actually read. It generates a detailed meeting report (key points, decisions, action items, open questions, themes), stores the output in a database for tracking, and pushes a structured version into Mem.ai so your knowledge base stays organized and searchable. Finally, it drafts a follow-up email in Outlook using the meeting context and participant info, keeping a human in the loop so you can review before sending.

The flow starts with a scheduled poll or a form submission, plus webhooks for the built-in web app view. Then it retrieves and merges transcript streams, runs AI summarization, and formats the results into HTML for email and clean content for Mem.ai. The output lands in three places: your knowledge base, your email drafts, and your internal tracking table.

What You Get: Automation vs. Results

Example: What This Looks Like

Say your team runs 10 Teams meetings a week. Manually, you might spend about 10 minutes finding the right transcript, 20 minutes summarizing it, and another 10 minutes drafting a follow-up email. That’s roughly 40 minutes per meeting, or about 7 hours a week. With this workflow, you trigger processing (or let it poll on schedule), wait for AI summarization, then skim a ready-to-send Outlook draft. Realistically, you’re down to about 5 minutes of review per meeting, so you get back around 6 hours most weeks.

What You’ll Need

  • n8n instance (try n8n Cloud free)
  • Self-hosting option if you prefer (Hostinger works well)
  • Microsoft 365 + Microsoft Graph to access Teams transcripts and meetings
  • Mem.ai for searchable meeting knowledge base storage
  • OpenAI API key (get it from the OpenAI API dashboard)

Skill level: Intermediate. You’ll connect accounts and handle Microsoft Azure app permissions (admin consent matters here).

Don’t want to set this up yourself? Talk to an automation expert (free 15-minute consultation).

How It Works

A scheduled check or a manual trigger starts it. The workflow can run on a schedule (polling recent meetings in a time window) and it also supports a form/webhook path for on-demand processing through a simple web view.

Transcript and meeting data gets retrieved and cleaned. n8n calls Microsoft Graph and OneDrive/SharePoint endpoints through HTTP Request nodes, expands lists of transcript files, and merges multiple streams into one coherent meeting record. Duplicate entries are removed so you don’t get double summaries.

AI turns the raw text into structured notes. The OpenAI chat model nodes and AI Agent nodes generate a summary plus the useful extras: action items, decisions, open questions, and themes. Markdown gets rendered into readable content, then it’s wrapped into a report you can reuse.

Results get stored and drafted for follow-up. Meeting records and outputs are stored in Postgres, Mem.ai receives the structured knowledge entry, and Outlook drafts are generated in HTML so you can review and send with confidence.

You can easily modify the summary format to match your company templates based on your needs. See the full implementation guide below for customization options.

Step-by-Step Implementation Guide

Step 1: Configure the Schedule and Webhook Triggers

Set up the triggers that start the workflow on a schedule or via incoming web requests.

  1. Add and configure Scheduled Poll Trigger to run on your preferred schedule.
  2. Configure Incoming Web Page as a webhook endpoint for rendering the web app response.
  3. Configure Next Step Webhook as the endpoint that continues processing after a user action.
  4. Set up Form Submission Trigger to accept form submissions and start the alternate intake path.

Tip: After saving each webhook node, copy the test URL and use it during setup of your web or form clients.

Step 2: Connect Postgres Data Sources

Configure the database read/write nodes used to fetch meeting records and store results.

  1. Open Verify Processing Status and connect the database used to check transcript processing state. Credential Required: Connect your Postgres credentials.
  2. Open Fetch Meeting Record and Fetch Meeting Record A to read meeting data for the web and follow-up flows. Credential Required: Connect your Postgres credentials.
  3. Open Store Results and connect your database for summary and transcript storage. Credential Required: Connect your Postgres credentials.

⚠️ Common Pitfall: Make sure your Postgres nodes return records even when no data is found; Verify Processing Status relies on downstream filtering.

Step 3: Configure Transcript and Meeting Retrieval

Set up the API retrieval and transformation pipeline that collects call logs, meeting info, and transcript files.

  1. Connect Assign Time Window to Retrieve Call Logs, then review Transform Logic A for your call-log parsing rules.
  2. Review Filter Join Link and Fetch Online Meeting Info to ensure join links and meeting metadata are properly filtered and expanded.
  3. Confirm the meeting list is expanded in Expand Meeting List, then deduplicated in Remove Duplicate Entries and ordered in Sort Records.
  4. Configure Fetch Transcript Data to collect transcript references, then expand each item in Expand Transcription List.
  5. Make sure Search OneDrive Meeting → Expand Items Batch → List Transcript Files → Retrieve Transcript B can access your storage API (add auth headers or OAuth2 as required).

⚠️ Common Pitfall: All httpRequest nodes in this flow call external APIs. Ensure each has the correct authentication (e.g., Microsoft Graph/OneDrive tokens) before testing.

Step 4: Normalize Transcript Subjects and Merge Streams

Align subject metadata from multiple transcript sources and merge them into a unified processing stream.

  1. Confirm Retrieve Transcript outputs into Map Transcript Subject and Retrieve Transcript B outputs into Map Transcript Subject B.
  2. Set any subject rules in Map Transcript Subject C based on the results of Conditional Check B.
  3. Merge all subject variations in Combine Streams, then pass them into Iterate Records for batch processing.

Tip: Conditional Check B routes to Map Transcript Subject C or Assign Time Window depending on your filtering logic.

Step 5: Set Up AI Summarization and Payload Creation

Configure the AI summary generation and the payload that drives email and storage outputs.

  1. Open OpenAI Chat Engine and add credentials. Credential Required: Connect your OpenAI credentials.
  2. Verify AI Generate Summary uses OpenAI Chat Engine as its language model (credentials are added to OpenAI Chat Engine, not the agent).
  3. Confirm AI Generate Summary outputs to Compose JSON Payload and that payload fields are mapped for downstream email and summary content.
  4. Compose JSON Payload outputs to both Build Email HTML and Build Summary HTML in parallel.

⚠️ Common Pitfall: If Iterate Records does not output data to its second output, AI Generate Summary will not run.

Step 6: Configure Email Delivery and Web Responses

Set up email rendering, Outlook delivery, and conditional web responses.

  1. Review Build Email HTML → Send Analysis Email and connect Microsoft Outlook. Credential Required: Connect your Microsoft Outlook credentials.
  2. Review Build Summary HTML → Store Results to store the generated summary content.
  3. Configure the web flow: Fetch Meeting Record → Render Markdown → Route by Condition.
  4. Ensure Route by Condition can send to Memo Request → Return Webhook Reply or to AI Draft Follow-up.
  5. Open OpenAI Chat Engine B and add credentials. Credential Required: Connect your OpenAI credentials.
  6. Verify AI Draft Follow-up uses OpenAI Chat Engine B as its language model (credentials are added to OpenAI Chat Engine B, not the agent).
  7. Configure Prepare Email JSON → Dispatch Follow-up Email → Return Status 200 and connect Microsoft Outlook. Credential Required: Connect your Microsoft Outlook credentials.

Step 7: Validate Processing Status and Matching Logic

Ensure the workflow only processes new or eligible transcripts and captures matches correctly.

  1. Confirm Verify Processing Status checks current records and passes outputs downstream.
  2. Verify Processing Status outputs to both Filter Empty Records and Retain Matches in parallel.
  3. Review Filter Empty Records → Collect Items → Retain Matches to ensure only valid items pass to Fetch Transcript Data.

Step 8: Test & Activate Your Workflow

Run end-to-end tests for each trigger path and activate the workflow for production.

  1. Use Scheduled Poll Trigger “Test” to ensure scheduled execution populates Retrieve Call Logs and downstream transcript processing.
  2. Send test payloads to Incoming Web Page and Next Step Webhook to confirm Return WebApp HTML and Return Webhook Reply respond correctly.
  3. Submit a test form to Form Submission Trigger and verify Transform Logic B → Combine Streams B processes as expected.
  4. Confirm successful runs send emails via Send Analysis Email and Dispatch Follow-up Email, and that summaries are stored by Store Results.
  5. When tests are successful, switch the workflow to Active for production use.
🔒

Unlock Full Step-by-Step Guide

Get the complete implementation guide + downloadable template

Common Gotchas

  • Microsoft Graph credentials can expire or need specific permissions. If things break, check your Azure App Registration API permissions and admin consent status first.
  • If you’re using Wait nodes or external rendering, processing times vary. Bump up the wait duration if downstream nodes fail on empty responses.
  • Default prompts in AI nodes are generic. Add your brand voice early or you’ll be editing outputs forever.

Related Workflows

If you already live in Outlook and want your day to start with less chaos, Outlook to Microsoft Teams, daily agenda posted pairs nicely with this setup by pushing the plan into Teams before the meetings even happen.

For teams that store recordings or exports in Drive first, Google Drive to Gmail, meeting recaps sent clean is a good companion workflow so recaps still go out even when the meeting wasn’t in Teams.

Need your meeting knowledge base to live somewhere more “wiki-like” than Mem.ai? When documentation is the priority, Google Meet to Notion, meeting notes everyone reads shows how to structure notes so they don’t get lost after publishing.

Agencies and distributed teams often want a simple default everyone can open. Zoom to Google Docs, meeting notes your team uses is useful when clients insist on Docs and you still want consistent summaries.

Finally, if your follow-ups turn into long Q&A threads, Outlook + Notion: answers stay organized and findable helps you keep decisions and clarifications searchable instead of buried in email chains.


Quick reference:

Frequently Asked Questions

How long does it take to set up this Teams Mem.ai automation?

Plan for about 1-2 hours, mostly for Microsoft permissions.

Do I need coding skills to automate Teams meeting notes into Mem.ai?

No. You will mostly be connecting accounts and copying settings from the setup guide. The hardest part is usually Azure admin consent, not writing code.

Is n8n free to use for this Teams Mem.ai automation workflow?

Yes. n8n has a free self-hosted option and a free trial on n8n Cloud. Cloud plans start at $20/month for higher volume. You’ll also need to factor in OpenAI API costs (often just a few dollars a month for normal meeting volume).

Where can I host n8n to run this automation?

Two options: n8n Cloud (managed, easiest setup) or self-hosting on a VPS. For self-hosting, Hostinger VPS is affordable and handles n8n well. Self-hosting gives you unlimited executions but requires basic server management.

Can I customize this Teams Mem.ai automation workflow for different note templates?

Yes, and you should. Most teams tweak the AI prompts in the “AI Generate Summary” and “AI Draft Follow-up” agent nodes so the output matches their tone and the sections they care about. You can also adjust the HTML in “Build Summary HTML” and “Build Email HTML” to match your internal format. Common tweaks include a stronger action-item table, tagging by project or client, and adding a “risks” section.

Why is my Microsoft Graph connection failing in this workflow?

Most of the time it’s missing admin consent or the wrong Graph scopes in your Azure App Registration. Regenerate the client secret if it expired, then re-authenticate the Generic OAuth2 credential in n8n. Also check that the meeting transcript actually exists for that call (some meetings don’t produce one), because the workflow will have nothing to retrieve in that case.

How many meetings can this Teams Mem.ai automation handle?

It depends on your hosting and how often you run it.

Is this Teams Mem.ai automation better than using Zapier or Make?

For this use case, usually yes. You’re pulling multiple files, merging streams, deduplicating records, and running multi-step AI generation, which gets awkward (and expensive) in simpler automation tools once you add branching and error handling. n8n also gives you the self-hosted option, which means you’re not paying per tiny step when volume grows. Zapier or Make can still be fine if you only want a basic “new transcript → summary → email” with minimal formatting. If you want help choosing, Talk to an automation expert and sanity-check the approach.

Once this is running, meeting notes stop being a recurring project. The workflow handles the repetitive cleanup, and you get decisions, action items, and follow-ups you can trust.

Need Help Setting This Up?

Our automation experts can build and customize this workflow for your specific needs. Free 15-minute consultation—no commitment required.

Talk to Automation Expert →

Lisa Granqvist

Workflow Automation Expert

Expert in workflow automation and no-code tools.

×

Use template

Get instant access to this n8n workflow Json file

Copy and import json workflow into your n8n workflow

💬
Get a free quote today!
Get a free quote today!

Tell us what you need and we'll get back to you within one working day.

Get a free quote today!
Get a free quote today!

Tell us what you need and we'll get back to you within one working day.