Skip to content

Audio Reports V2

An intelligent WhatsApp-based reporting system that processes voice messages and text inputs from users in educational programs, providing AI-powered responses and generating structured reports based on user roles and program participation.

Purpose

No business context provided yet — add a context.md to enrich this documentation.

This workflow appears to serve educational programs (EBA and EXP) by: - Onboarding new users and assigning them appropriate roles - Processing voice messages and text inputs via WhatsApp - Generating AI-powered responses and reports based on user context - Maintaining user data and conversation history in Google Sheets

How It Works

  1. Message Reception: The workflow receives WhatsApp messages (text or voice) through a webhook
  2. Voice Processing: If the message is a voice note, it downloads and transcribes the audio using OpenAI
  3. User Validation: Checks if the sender exists in the user database and has complete profile information
  4. User Onboarding: For new or incomplete users, runs an onboarding process to collect name and assign role (EBA/EXP program participant, trainer, or mentor)
  5. Role Assignment: Updates the user database with collected information and assigned role ID
  6. AI Processing: For existing users, retrieves role-specific prompts and processes the message through an AI agent
  7. Report Generation: Creates structured reports in markdown format based on the user's input and role
  8. Response Delivery: Sends AI-generated responses back to the user via WhatsApp
  9. Data Storage: Saves generated reports and timestamps to the Google Sheets database

Workflow Diagram

graph TD
    A[WhatsApp Webhook] --> B{Voice Note?}
    B -->|Yes| C[Download Voice File]
    B -->|No| D[Text Input]
    C --> E[Transcribe Audio]
    E --> F[Voice Output]
    D --> F
    F --> G[User Message]
    G --> H{User Exists?}
    H -->|Yes| I[Check User in DB]
    I --> J{Complete Profile?}
    J -->|Yes| K[Get Current Prompt]
    J -->|No| L[Onboarding Agent]
    L --> M[Assign Role]
    M --> N[Send Response]
    K --> O[AI Agent]
    O --> P[Upsert Report]
    P --> Q[Send Response]

Trigger

Webhook: POST endpoint at path 23d0efd9-c4dc-4cdd-82a8-f389082afdf2 that receives WhatsApp messages from Twilio.

Nodes Used

Node Type Purpose
Webhook Receives incoming WhatsApp messages from Twilio
HTTP Request Downloads voice message files from WhatsApp
OpenAI (Transcription) Converts voice messages to text
Google Sheets Reads/writes user data, prompts, and reports
If (Conditional) Routes workflow based on message type and user status
Set Formats and structures data between nodes
AI Agent (LangChain) Processes messages with role-specific AI prompts
Structured Output Parser Ensures AI responses follow defined JSON schemas
OpenAI Chat Model Provides language model capabilities for AI agents
Memory Buffer Maintains conversation context per user
Twilio Sends responses back to users via WhatsApp

External Services & Credentials Required

Google Sheets API

  • Service Account: For reading/writing user data, prompts, and reports
  • Document ID: 1Vv_gTpvI8RHPPrn0iwzypG3qKRYeLDb1HZFPDzYe5ng (E Lab Report spreadsheet)
  • Sheets: Users, VoiceNotePrompts

OpenAI API

  • API Key: For audio transcription and chat completions
  • Models Used: GPT-4.1, GPT-4.1-mini

Twilio API

  • Account SID & Auth Token: For WhatsApp messaging
  • WhatsApp Number: +14155238886

Environment Variables

No explicit environment variables are defined in this workflow. All configuration is handled through n8n credentials and node parameters.

Data Flow

Input

  • WhatsApp messages (text or voice) containing:
    • User phone number (WaId)
    • Message content or audio file URL
    • Message metadata

Processing

  • User profile lookup and validation
  • Voice-to-text transcription (if applicable)
  • Role-based AI prompt selection
  • Conversation memory management
  • Structured response generation

Output

  • WhatsApp responses to users
  • Updated user records in Google Sheets
  • Generated reports stored with timestamps
  • Conversation history maintained in memory

Error Handling

The workflow includes several conditional paths for error handling:

  • Unknown Users: Redirects to onboarding flow for profile completion
  • Incomplete Profiles: Triggers role assignment process
  • Message Type Detection: Handles both voice and text inputs appropriately
  • Always Output Data: User existence check node configured to continue workflow even if no matches found

Known Limitations

  • Workflow is currently archived and inactive
  • Response sending nodes are disabled
  • No explicit error handling for API failures or malformed data
  • Memory buffer is session-based but may not persist across workflow restarts

No related workflows mentioned in the provided context.

Setup Instructions

  1. Import Workflow: Import the JSON configuration into your n8n instance

  2. Configure Credentials:

    • Set up Google Service Account with access to the target spreadsheet
    • Configure OpenAI API credentials
    • Set up Twilio account with WhatsApp sandbox or approved number

  3. Update Webhook URL:

    • Activate the webhook node to get the production URL
    • Configure Twilio to send WhatsApp messages to this endpoint

  4. Prepare Google Sheets:

    • Ensure the spreadsheet contains "Users" and "VoiceNotePrompts" sheets
    • Set up appropriate column headers matching the node configurations

  5. Test Configuration:

    • Send a test message to verify webhook reception
    • Check user onboarding flow with a new phone number
    • Verify AI responses and report generation

  6. Activate Workflow:

    • Enable the workflow in n8n
    • Re-enable the Twilio response nodes if needed
    • Monitor execution logs for any issues