Automated Google Meet Summary and Slack Notification - n8n Workflow

Teams using Google Meet extensively who need instant meeting summaries.
Technical users looking for robust n8n templates integrating multiple AI and communication services.
Project managers requiring automated action item tracking via Slack.
Developers exploring advanced language processing features within an n8n workflow.

This powerful n8n workflow solves the problem of manual meeting documentation. By integrating the Google Calendar n8n trigger, the Vexa.ai transcription service, and the latest GPT-4o model via specific n8n nodes, it creates a fully automated, end-to-end pipeline. When a meeting begins, the n8n trigger starts the process, ensuring a bot joins to record. Once the transcription is complete, the data is pre-processed and then fed into the AI Agent n8n node. This agent uses a specialized prompt to generate highly concise, actionable summaries for immediate distribution in Slack. This specific n8n workflow saves significant time, ensures key decisions are not lost, and provides a powerful example of real-world AI automation using n8n templates.

1. Event Start: The core n8n trigger (Google Calendar Trigger node) initiates the execution when a scheduled event (with a Google Meet link) begins.


2. Bot Deployment: An HTTP Request n8n node (Add bot to meet) sends the meeting ID extracted from the calendar event to the Vexa.ai API, instructing their service to deploy a recording bot to the live session.


3. Transcript Retrieval: The n8n workflow attempts to retrieve the transcript using the Get Vexa Transcript HTTP Request n8n node.


4. Polling Logic: A crucial If n8n node checks if the transcript status returned is 'completed'. If the transcript is not yet ready, the flow is routed to the Wait n8n node, creating a robust polling loop that periodically retries the retrieval request until Vexa.ai confirms completion.


5. Data Pre-processing: A specialized Code n8n node processes the raw, segmented transcript data from Vexa.ai. It concatenates the segments into a clean fullTranscript string, extracts unique participant lists, and calculates the meeting duration, optimizing the input for the AI.


6. AI Summarization: The cleaned data is passed to the AI Agent n8n node, which utilizes the connected OpenAI Chat Model (specifically GPT-4o) to execute a structured summary prompt. This ensures the output is highly focused on key decisions and action items, tailored for a Slack message.


7. Notification: Finally, the Send a message Slack n8n node publishes the concise, GPT-4o-generated summary directly to a specified Slack channel, completing the automated n8n workflow.

1. Import the n8n Workflow: Download the provided n8n workflow JSON and import it directly into your n8n instance.


2. Set Up Credentials: This n8n template requires four distinct credentials:

1. Review Configuration: Configure the Google Calendar Trigger to monitor the correct calendar ID. Verify the Send a message n8n node is set to the correct Slack channel name (e.g., #general).


2. Activate: Save the n8n workflow and activate the primary n8n trigger to begin monitoring for meeting starts.

Google Calendar Trigger: The starting n8n trigger for the entire flow. It monitors your calendar and executes the n8n workflow only when a scheduled event with a Google Meet link begins.
Add bot to meet (HTTP Request): Sends a POST request to Vexa.ai, passing the Google Meet conference ID to instruct the transcription bot to join the call.
Get Vexa Transcript (HTTP Request): Polls Vexa.ai for the final, processed transcript data. This n8n node is part of the retry logic.
If n8n node: A core flow control mechanism. It checks if the transcript status is completed. If not, it routes the execution through the Wait n8n node to retry the transcription request.
Wait n8n node: Pauses the n8n workflow execution flow control within the polling loop, preventing excessive API requests while waiting for the transcript to be ready.
Code n8n node: Performs essential data transformation. It aggregates segmented transcript data into a single string, extracts participant names, and calculates meeting duration, preparing the data for the next AI n8n node.
OpenAI Chat Model: This n8n node specifies the language model used—chatgpt-4o-latest—providing the powerful language processing backbone for the summarization task.
AI Agent n8n node: This critical component applies the detailed, structured prompt to the transcript, leveraging the linked OpenAI model to generate a concise, Slack-ready summary.

• Send a message (Slack): The final n8n node in the execution path, responsible for publishing the formatted, AI-generated summary to the designated team Slack channel.