PFNApp
Documentation
EnglishBahasa Indonesia
Console
Documentation/WhatsApp
Verified GuideWhatsAppEN

Visual Canvas & WhatsApp AI Bot Workflows

Comprehensive guide for visually designing, testing with live simulation, and deploying automated WhatsApp chatbots and AI agent workflows.

3 min read•Last updated 9/14/2026•/whatsapp/workflows

The AI & Bot Builder (/console/whatsapp/workflows) empowers organizations to visually build multi-step conversational bots, automated customer support triage, dynamic catalog lookups, and AI-driven sales agents without managing backend server logic.

Workflows Dashboard

1. Workflows Overview & Management

In the AI & Bot Workflows overview:

  • Active Flows: Monitor all deployed automations, node counts, and assigned phone numbers.
  • Trigger Indicators: View trigger conditions (e.g. whatsapp_inbound on +6283138855774).
  • Create Workflow: Start a clean workflow canvas from scratch or clone an existing template.
  • Open Canvas: Enter the interactive drag-and-drop flow builder.

2. Interactive Visual Canvas Builder

Click "Open canvas" (/console/whatsapp/workflows/[id]/canvas) to access the full-screen visual workspace.

Visual Workflow Canvas

Canvas Controls & Toolbar:

  • Workflow Name & Status: Edit automation name and inspect deployment state (Live / Draft).
  • ADD A STEP Toolbar:
    • AI Assist: Generate nodes and logic automatically with AI Copilot.
    • Send message: Dispatch formatted text or media bubbles to the customer.
    • Ask for input: Prompt customer for text or numbers and store the answer in a custom variable (e.g. customer_need).
    • Condition: Branch execution based on variable evaluation.
    • Interactive buttons: Provide quick-reply buttons or list messages.
    • AI response: Run an LLM prompt using conversation context and external API results.
    • HTTP request: Perform real-time GET/POST requests against external endpoints or internal catalog APIs.
  • Deletable Edges: Connect any output port to an input port. Hover or click an edge to reveal the Delete edge action.
  • Viewport Navigation: Pan, zoom in/out, and click Fit View for complete flow overview.

Dynamic Variable Interpolation:

Nodes pass data down the chain using clean template tags:

  • {{variables.<variable_name>}}: Values captured from Ask for input steps.
  • {{steps.<node_id>.body}}: JSON response bodies returned by HTTP request nodes.

3. Interactive Bot Simulator

Before deploying changes to live production devices, test your logic end-to-end using the built-in simulator.

Click "Simulate test" in the canvas header to open the simulator dialog:

WhatsApp Bot Simulator

  • Realistic Conversation Preview: See bot responses exactly as they will render on customer mobile devices.
  • Interactive Chat Input: Type simulated customer queries to verify AI prompt reasoning and HTTP API integrations.
  • Reset Session: Wipe the simulator memory to test fresh onboarding flows from step 1.

4. Deploying to Live Production

Once validated in the simulator:

  1. Click "Save and deploy" in the top navigation.
  2. The workflow will compile and activate immediately for all inbound messages on the assigned WhatsApp device.