PayPilot Backend

PayPilot is an AI-powered invoice generator that integrates directly with WhatsApp. It allows users to create, view, and manage invoices using natural language text or voice notes.

The system uses Groq (Llama 3) for intent parsing, Sarvam AI for voice-to-text transcription, and Puppeteer for generating professional PDF invoices. All data is stored in Supabase.

The standard government invoice format is used in pdf generation.

Youtube link

https://youtu.be/tMPPqJDV8sE

Features

• WhatsApp Interface: Interact with the bot entirely through WhatsApp.
• Natural Language Invoicing: "Create invoice for 5000 to ABC Corp for consulting"
• Voice Support: Send voice notes (English/Hindi), powered by Sarvam AI.
• Smart Intent Parsing: Uses LLMs (Groq/Llama 3) to extract structured data from text.
• PDF Generation: Automatically generates and sends PDF invoices.
• Payment Tracking: Mark invoices as paid using natural language.
• Scheduler: Automated checks for overdue invoices.

Tech Stack

• Runtime: Node.js
• Framework: Express.js
• Database: Supabase (PostgreSQL)
• WhatsApp Integration: whatsapp-web.js
• AI / LLM: Groq SDK (Llama 3.3 70B Versatile)
• Speech-to-Text: Sarvam AI
• PDF Generation: Puppeteer
• Scheduling: Node-cron

Project Structure

backend/
├── src/
│   ├── services/
│   │   ├── intentParser.js    # LLM interaction (Groq)
│   │   ├── invoiceService.js  # DB operations (Supabase)
│   │   ├── pdfService.js      # PDF generation (Puppeteer)
│   │   ├── schedulerService.js# Cron jobs for reminders
│   │   ├── sttService.js      # Speech-to-text (Sarvam AI)
│   │   ├── supabaseClient.js  # DB connection
│   │   └── whatsappService.js # WhatsApp bot logic
│   ├── index.js               # Enry point
│   └── setupDatabase.js       # Database initialization script
├── .env                       # Environment variables
├── package.json               # Dependencies
└── README.md                  # Documentation

Setup & Installation

1. Clone the Repository

git clone <repository-url>
cd PayPilot/backend

2. Install Dependencies

npm install

3. Environment Variables

Create a .env file in the backend/ directory with the following keys:

PORT=3000

# WhatsApp (Optional, handled by whatsapp-web.js)
# WHATSAPP_CLIENT_ID=

# Supabase Configuration
SUPABASE_URL=your_supabase_url
SUPABASE_ANON_KEY=your_supabase_anon_key

# AI Services
GROQ_API_KEY=your_groq_api_key
SARVAM_API_KEY=your_sarvam_api_key

4. Database Setup

Run the setup script to create the necessary tables in your Supabase project:

npm run setup-db

Note: This script creates an invoices table if it doesn't exist.

5. Start the Server

npm start

OR for development with auto-reload:

npm run dev

How to Use

1. Start the Bot: Run the server. It will generate a QR code in the terminal.
2. Scan QR Code: Use WhatsApp on your phone to scan the QR code (Linked Devices explained below).
3. Send Commands:
   • Create Invoice:

     "Create invoice of ₹15,000 to ABC Traders for consulting, due in 7 days"

   • Check Status:

     "Show invoice status for #1234"

   • Mark as Paid:

     "Mark invoice #1234 as paid"

   • Voice Note:

     Record a voice note saying the same commands.

╔══════════════════════════════════════════════════╗
                WHATSAPP QR CODE — SCAN ME!                           
╠══════════════════════════════════════════════════╣
  How to connect:                                                
  1. Open WhatsApp on your phone                                 
  2. Tap ⋮ (Menu) → "Linked Devices"                               
  3. Tap "Link a Device"                                         
  4. Point your phone camera at the QR below                     
╚══════════════════════════════════════════════════╝

Screenshots

QR code to link your mobile to the system through whatsapp

Transcribed text

Pdf auto generated