Jayson Bautista.

WhatsApp content approval.

A conversational content flow with caption generation and a separate file-save route.

  • n8n
  • WhatsApp
  • Gemini
  • Google Sheets
  • Google Drive

Personal project · Source-backed architecture

The problem explored

Separating content generation from a file-save route inside a WhatsApp conversation.

A conversational content flow with caption generation and a separate file-save route. The supplied diagrams document the configured connections, rather than verified production outcomes.

Authentic cropped source architecture from my WhatsApp content approval n8n project
WhatsApp content approvalEditor whitespace removed; original nodes preserved
Full source

Scroll sideways to inspect the diagram. Zoom enlarges labels; Reset zoom restores this preview.

Architecture summary · read alongside the source

  1. Receive an instruction
  2. Choose a route
  3. Save content
  4. Generate a caption

Inside the processing logic

Receive an instruction

  • WhatsApp Trigger starts the workflow.
  • An AI Agent Parser connects to Gemini.
  • A Parser code node prepares data before an If node.

Choose a route

  • The If node separates two visible branches.
  • One branch reads rows from Google Sheets before a code node.
  • The condition determining the route is not visible.

Save content

  • The upper branch uses Google Drive Create File from Text.
  • A WhatsApp send-message node follows the file step.
  • The screenshot does not expose folder configuration or approval wording.

Generate a caption

  • The lower branch reads Google Sheets and uses an AI Agent Caption.
  • Parser and Code for WhatsApp nodes prepare the outgoing message.
  • WhatsApp delivery is followed by an Append Row step in Google Sheets.

Step through the explanation

Dennice, an original illustrated robot guide, walks through the diagram. Start with Next, or choose autoplay. Captions and a full transcript are included; sound is optional.

Open the film on its own

Source evidence

These are the supplied local files used to write this explanation. Node labels and connections are visible; exported workflow JSON and execution logs were not supplied.

WhatsApp content approval source diagram 1
Source: automation8/1.png. Open original image

Visible outputs

A generated caption or a text file saved to Google Drive, with WhatsApp feedback.

These outputs are represented by nodes in the source. Delivery success and business results require execution evidence.

What this case study does—and doesn't—verify

The diagram shows generation and file-save routes. The number of captions and exact approval commands are not evidenced. No publishing integration or measured engagement result is claimed.

Next validation steps

Before operational use, I would inspect the workflow export, confirm credentials and permissions, test representative inputs and error paths, and review the outgoing messages with the intended user. These are proposed checks, not completed workflow execution tests.

Next: AI email management

Let’s make
something work.

Have a development role, an automation idea or a website project in mind? I’d like to hear about it.

Jayson Bautista · Automation Developer & Web Developer