Skip to content

AudioReportsV2

An AI-powered WhatsApp assessment workflow that guides trainers through structured learner evaluations using voice notes and generates comprehensive assessment reports for skill-up program participants.

Purpose

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

This workflow appears to be designed for educational or training programs where mentors/trainers need to conduct structured assessments of learners. It facilitates the assessment process through WhatsApp by providing scenario-based evaluation frameworks, collecting voice note feedback, and generating detailed reports on learner progress.

How It Works

  1. Assessment Initiation: Trainer sends "start" message to begin the assessment process
  2. Scenario Generation: AI generates three learner scenarios (Struggling, Progressing, Excelling) for the trainer to choose from
  3. Scenario Selection: Trainer selects a scenario (A, B, or C) that best matches their learner
  4. Question Generation: AI creates targeted assessment questions based on the selected scenario
  5. Voice Response Collection: Trainer responds with voice notes answering the assessment questions
  6. AI Follow-up: System transcribes voice notes and provides intelligent follow-up questions
  7. Assessment Storage: All responses are stored and tracked across multiple assessment areas
  8. Report Generation: When trainer requests "generate report", AI creates a comprehensive assessment report
  9. Report Delivery: Final assessment report is sent via WhatsApp in formatted sections

Workflow Diagram

graph TD
    A[WhatsApp Webhook] --> B[Check Start Message]
    A --> C[Check Selection]
    A --> D[Check Response]
    A --> E[Check Report Request]

    B --> F[Generate Scenarios]
    F --> G[Store Scenarios]
    G --> H[Send Scenarios]

    C --> I[Process Selection]
    I --> J[Generate Questions]
    J --> K[Send Questions]

    D --> L[Prepare Voice Note]
    L --> M[Transcribe Voice Note]
    M --> N[Generate AI Follow-up]
    N --> O[Store Transcription]
    O --> P[Update Assessment Memory]
    P --> Q[Send AI Follow-up]

    E --> R[Generate Report]
    R --> S[Send Report]

Trigger

WhatsApp Webhook - Listens for incoming WhatsApp messages on the /coaching-start endpoint. The workflow responds to different message types: - Messages containing "start" to begin assessments - Single letter responses (A, B, C) for scenario selection - Voice notes for assessment responses - Messages containing "generate report" for final report creation

Nodes Used

Node Type Purpose
Webhook Receives incoming WhatsApp messages
If (Conditional) Routes messages based on content (start, selection, voice notes, report requests)
OpenAI Generates scenarios, questions, follow-ups, and final reports using GPT
Code Processes selections, prepares voice notes, and manages assessment data storage
Set Stores scenario data and updates assessment memory
Twilio Sends WhatsApp messages back to users

External Services & Credentials Required

OpenAI API

  • Purpose: Text generation, voice transcription
  • Credential ID: E8hkxFY5Lqqn03I5
  • Required: OpenAI API key with access to chat completions and audio transcription

Twilio API

  • Purpose: WhatsApp messaging
  • Credential ID: 5owj9iWkYcotiAh4
  • Required: Twilio Account SID, Auth Token, and configured 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 and voice notes)
  • Message metadata (sender, timestamp, media URLs)

Processing

  • AI-generated assessment scenarios and questions
  • Voice note transcriptions
  • Structured assessment data storage
  • Progressive conversation memory

Output

  • Interactive WhatsApp conversations
  • Structured assessment reports
  • Learner evaluation summaries with strengths, development areas, and recommendations

Error Handling

The workflow includes several error handling mechanisms:

  • Voice Note Validation: Checks for media presence and valid URLs before processing
  • Scenario Parsing: Fallback to default scenarios if AI-generated content is malformed
  • Data Storage: Safe handling of missing assessment data with default values
  • Selection Validation: Regex validation for scenario selection (A, B, C only)

Known Limitations

  • WhatsApp message length is limited to 1600 characters per message
  • Voice note processing requires valid media URLs from Twilio
  • Assessment memory is stored in workflow execution context (not persistent across workflow restarts)
  • Currently configured for a single WhatsApp number (+14155238886)
  • Workflow is currently disabled (active: false)

No related workflows are mentioned in the current configuration.

Setup Instructions

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

  2. Configure Credentials:

    • Set up OpenAI API credentials with chat and audio transcription access
    • Configure Twilio credentials with WhatsApp messaging enabled

  3. Webhook Configuration:

    • Enable the webhook trigger
    • Configure Twilio to send WhatsApp messages to the webhook URL
    • Ensure the webhook path matches /coaching-start

  4. Test the Flow:

    • Send "start" to the configured WhatsApp number
    • Follow the assessment flow with scenario selection
    • Test voice note responses and report generation

  5. Activate Workflow: Set the workflow status to active to begin processing messages

  6. Monitor Performance: Check execution logs for any API rate limits or transcription errors