Video Call Agent Analytics and Reporting to Google Sheets - n8n Workflow

• Businesses utilizing AI video agents (like BeyondPresence) for sales or support calls.


• Data analysts needing structured, real-time reporting on conversational data.


• Operations teams looking for automated sentiment analysis and action item extraction.


• Users searching for robust n8n templates to handle webhooks and large language models.

Managing high volumes of video agent calls requires efficient, structured reporting. This powerful n8n workflow solves the manual burden of call review by automatically capturing, analyzing, and structuring conversational data.

Triggered immediately upon call completion via a webhook, the workflow uses an n8n node configured for GPT-4o-mini to perform deep conversational analysis. It extracts key metrics like user sentiment, action items, and discussion points, enforcing a strict JSON output format. Finally, this structured output is seamlessly appended to a Google Sheets spreadsheet, transforming raw transcripts into actionable, searchable data points for immediate business intelligence. This specific n8n workflow is designed for scalability and reliability.

This highly efficient n8n workflow follows seven distinct steps:

1. Trigger Reception: The process begins when the BeyondPresence Webhook n8n trigger receives a POST request containing the full call transcript and metadata after a call concludes.


2. Immediate Acknowledgment: The Acknowledge Webhook n8n node immediately responds with a 200 status code, ensuring the sending system doesn't experience delays.


3. Event Filtering: The Filter Call Events n8n node ensures that processing only continues if the event type is specifically call_ended, preventing processing of non-essential triggers.


4. Data Validation and Enrichment: A Code n8n node validates the integrity of the incoming data, calculates crucial fields (like duration in minutes, time of day, and checks for preliminary action items), enriching the payload before AI analysis.


5. AI Analysis: The AI Call Analysis n8n node (configured with a powerful large language model like GPT-4o-mini) receives the cleaned data and uses a detailed system prompt to analyze the entire conversation, generating a comprehensive summary, including key decisions and action items, strictly formatted as JSON.


6. JSON Parsing: The Parse AI Response Code n8n node handles common LLM formatting inconsistencies, carefully extracting the clean JSON object from the AI's text response, often stripping markdown delimiters and repairing common errors.


7. Reporting: The final Save to Google Sheets n8n node takes the structured, analyzed data and maps it to specific columns (Title, Participant, Action Item 1, Sentiment Overall, etc.) appending a new, comprehensive row to the designated Google Sheet, completing the full n8n workflow cycle.

To deploy this comprehensive n8n workflow template, follow these steps:

1. Import: Copy the provided n8n workflow JSON and paste it into your n8n instance under the 'Workflows' section.


2. BeyondPresence Setup (n8n trigger): Activate the BeyondPresence Webhook n8n trigger node. Copy the unique Test or Production Webhook URL generated by this n8n node. Navigate to your BeyondPresence dashboard settings and configure this URL as your Call End Webhook endpoint.


3. OpenAI Credential Setup: Update the 'AI Call Analysis' n8n node with your OpenAI or OpenRouter credentials.


4. Google Sheets Preparation: Use the provided link to copy the required Google Sheets template. Obtain the new Spreadsheet ID from your browser's URL bar.


5. Google Sheets Credential Setup: Configure the 'Save to Google Sheets' n8n node. Add your Google Sheets credentials and paste the copied Spreadsheet ID into the document ID field.


6. Activation: Ensure all credentials are set and save the n8n workflow. Set the workflow to 'Active' to start listening for call triggers.

BeyondPresence Webhook (n8n trigger):
Function: Acts as the primary n8n trigger, initiating the automation when a video call ends.
Key Configuration: Path is set to beyondpresence-call-webhook, configured to accept POST requests.
Acknowledge Webhook (Respond To Webhook n8n node):
Function: Sends an immediate 200 response to the webhook sender to prevent timeouts.
Key Configuration: Responds with a simple JSON message: {"status": "success", ...}.
Validate & Enrich Data (Code n8n node):
Function: Performs critical data hygiene, validating fields, and calculating derived values like durationMinutes, dayOfWeek, and checking for action keywords, ensuring the AI receives optimal input.
AI Call Analysis (Langchain OpenAI n8n node):
Function: The core analytical engine. It uses the gpt-4o-mini model to transform the raw transcript into a highly structured JSON report based on detailed instructions.
Key Configuration: Uses custom system and user prompts to mandate specific outputs like metadata, actionitems, and sentimentanalysis in a rigid JSON format.
Parse AI Response (Code n8n node):
Function: Robustly extracts the clean JSON object from the text output of the language model, handling common issues like markdown code blocks ( `json ) and trailing commas.
Save to Google Sheets (Google Sheets n8n node):
Function: The final action, persisting the analyzed data into a centralized spreadsheet.
Key Configuration: Operation: append. Maps dozens of fields extracted from the AI's JSON output (e.g., {{ $json.metadata.topic }}, {{ $json.action_items[0]?.task }}) to corresponding sheet columns.