Ajay – WhatsApp AI Chatbot with Personality

Ajay is a WhatsApp-based AI chatbot that interacts with users in real-time using the WhatsApp Cloud API.
Built as a fun and creative project, Ajay embodies a flamboyant, Bollywood-style persona.

The bot responds to messages with charm, wit and a dramatic personality — all powered by AI.

---

Features

• Real-time WhatsApp chat
• Custom AI persona (Ajay)
  • Hindi + English mix
  • Confident, flirtatious, dramatic tone
  • Playful jealousy and humor
• Instant replies via webhook
• AI-powered responses using OpenAI
• Deployed on Railway (24/7 availability)
• Secure environment-based configuration

---

How It Works

1. A user sends a message to the WhatsApp number
2. WhatsApp Cloud API forwards it to a webhook
3. FastAPI server processes the message
4. The message is sent to OpenAI with a custom persona prompt
5. AI generates a response
6. Response is sent back to the user via WhatsApp API

---

Tech Stack

• Backend: FastAPI (Python)
• Messaging: WhatsApp Cloud API (Meta)
• AI: OpenAI API (gpt-4.1-mini)
• Hosting: Railway
• Tunneling (dev): ngrok
• Database (optional): SQLite / PostgreSQL
• Environment Management: .env

---

Project Structure

.
├── main.py          # FastAPI app + webhook handling
├── brain.py         # AI response generation logic
├── persona.py       # Ajay's personality definition
├── db.py            # (optional) chat logging
├── requirements.txt
├── Procfile
└── README.md

---

Setup Instructions

1. Clone the Repository

git clone https://github.com/your-username/ajay-whatsapp-bot.git
cd ajay-whatsapp-bot

2. Install Dependencies

pip install -r requirements.txt

3. Create .env file

WHATSAPP_TOKEN=your_meta_permanent_token
PHONE_NUMBER_ID=your_phone_number_id
VERIFY_TOKEN=your_custom_verify_token
OPENAI_API_KEY=your_openai_api_key

4. Run Locally

uvicorn main:app --reload --port 8000

5. Expose via ngrok (for development)

ngrok http 8000

Use the generated URL for webhook configuration:

https://<ngrok-id>.ngrok-free.app/webhook

6. Configure Meta Webhook

• Go to Meta Developer Dashboard
• Add webhook URL
• Set verify token
• Subscribe to:
  • messages

7. Deploy on Railway (Production)

1. Push code to GitHub
2. Go to Railway → New Project → Deploy from GitHub
3. Add environment variables in Railway dashboard
4. Replace webhook URL with Railway URL:

https://your-app.up.railway.app/webhook

---

Environment Variables

Variable	Description
WHATSAPP_TOKEN	Permanent access token from Meta
PHONE_NUMBER_ID	WhatsApp Business phone number ID
VERIFY_TOKEN	Custom token for webhook verification
OPENAI_API_KEY	OpenAI API key

---

Cost Notes

• OpenAI usage is extremely low (₹5–₹50 per week for moderate usage)
• Railway free credits (~$5) are sufficient for short-term usage
• Suitable for small-scale or personal projects

---

Limitations

• WhatsApp requires a 24-hour conversation window
• AI replies depend on OpenAI quota
• SQLite (if used) is not persistent on Railway without volumes

---

Use Cases

• Fun chatbot projects
• Experimenting with AI personalities
• Learning WhatsApp Cloud API integration
• Event-based interactive experiences

---

Future Improvements

• Persistent memory (PostgreSQL)
• Admin dashboard to monitor chats
• Voice note responses
• Scheduled messages
• Dynamic persona control

---