AutomationsTeam ManagementAI ToolsReporting

How to Auto-Generate Weekly Team Reports Using Google Meet Transcript + Claude API

4 min read

Automate your weekly team reporting. Use Google Meet transcripts and Claude API to generate concise, actionable weekly reports automatically.

The "Weekly Sync" is a necessary evil.

You need it to align the team. But the aftermath—writing up the notes, assigning the tasks, and summarizing the wins for stakeholders—is a chore that often gets skipped.

When notes get skipped, accountability dies. "Oh, I thought we said we'd do that next week."

The Solution: Let the robots listen, write, and report.

This workflow takes the raw audio/transcript from your Google Meet, feeds it to Claude (Anthropic's AI), and automatically emails a perfectly formatted "Weekly Executive Summary" to your team.

The Workflow Overview

  • Record: You hold your meeting on Google Meet (or Zoom).
  • Trigger: The transcript file is saved to Google Drive.
  • Analyze: Claude reads the entire 45-minute conversation.
  • Extract: It pulls out:
  • * ✅ Decisions Made

    * 🚧 Blockers

    * 📅 Action Items (with owners)

  • Deliver: An email summary is sent to the team.
  • Step 1: Getting the Transcript

    Google Meet has a built-in "Transcripts" feature (if you have a Workspace plan). When you turn it on, it saves a Google Doc to a specific folder in Drive called "Meet Recordings".

    Make.com Trigger:

    * Module: Google Drive > Watch Files in Folder.

    * Folder: Select your "Meet Recordings" folder.

    * Filter: Only watch files with "Transcript" in the name.

    Step 2: The AI Analyst (Claude)

    We prefer Claude 3.5 Sonnet for this task because it writes in a very natural, professional tone and handles large amounts of text effortlessly.

    * Module: Anthropic (Claude) > Create a Message.

    * Input: The text content of the Google Doc.

    The Prompt:

    > "You are a Chief of Staff. I am providing a transcript of a weekly team meeting.

    >

    > Please generate a 'Weekly Status Report' with the following sections:

    > 1. Executive Summary: 2-3 sentences on the overall vibe and main focus.

    > 2. Key Decisions: Bullet points of what was agreed upon.

    > 3. Blockers: Anything stopping progress.

    > 4. Action Items: A table of Task | Owner | Deadline (if mentioned).

    >

    > Tone: Professional, concise, direct. Ignore small talk."

    Seeing this in action makes a huge difference. Once configured, the interface provides a clear visual confirmation that your automation is running smoothly, giving you peace of mind that the system is working as intended.

    Step 3: The Delivery (Email or Slack)

    Now you have the text. Where should it go?

    Option A: Email (Gmail Module)

    * To: team@yourcompany.com

    * Subject: Weekly Report - {{MM/DD/YYYY}}

    * Body: The output from Claude.

    Option B: Slack (Slack Module)

    * Channel: #team-updates

    * Message: "📢 Weekly Sync Summary is ready!\n\n" + Claude Output.

    Advanced: Auto-Assigning Tasks

    If you want to be a productivity god, add a step to parse the "Action Items".

    Ask Claude to output the action items as JSON instead of text.

    ``json

    [

    {"task": "Fix the login bug", "owner": "Sarah", "due": "Friday"},

    {"task": "Email the client", "owner": "Mike", "due": "Tomorrow"}

    ]

    ``

    Then, use an Iterator in Make.com to loop through these items and create tasks in Notion or Asana.

    Result:* You finish the meeting, and 5 minutes later, Sarah gets a notification in Asana: "New Task: Fix the login bug."

    This small adjustment can have a significant impact on your overall workflow efficiency. It turns a manual, error-prone process into a reliable, set-it-and-forget-it system.

    Why Claude vs. GPT-4?

    For creative writing, they are similar. For summarization, we find Claude tends to be less "fluffy." It gets straight to the point. It also has a massive context window (200k tokens), so you can feed it a 3-hour quarterly planning meeting and it won't crash.

    Conclusion

    This automation turns a passive activity (sitting in a meeting) into an active documentation process. You build a searchable history of every decision your company has ever made, without lifting a finger.

    Next Step: Enable "Transcripts" in your Google Meet settings today. That's the first step.

    * Start automating with Make.com


    Want to summarize 1-on-1 client calls instead? Check out our guide on Auto-Summarizing Calls with Fireflies.

    Frequently Asked Questions

    Why use Claude instead of ChatGPT?

    Claude (specifically Claude 3.5 Sonnet) is widely considered superior for summarization tasks because it has a larger context window (can read longer transcripts) and tends to hallucinate less when extracting specific details like deadlines.

    Does this work with Zoom?

    Yes. If you use Zoom's native cloud recording, you can trigger the automation when a new recording is available. The logic is identical: Get Transcript -> Send to AI -> Email Report.

    Is it private?

    You are sending your meeting data to the Anthropic API. Their enterprise terms generally state they don't train on API data, but you should always check your specific agreement. Avoid using this for highly sensitive legal or medical discussions.

    Can it assign tasks in my project management tool?

    Yes! In the 'Action Items' step, you can have Make.com iterate through the list and create tasks in Asana, ClickUp, or Trello automatically. That's an advanced version of this workflow.

    Share this article

    Want more tips like this?

    Explore our blog for more guides on AI automation for solopreneurs.

    Browse All Articles