Movie Character Chatbot

Overview

The Movie Character Chatbot is an AI-powered chatbot that allows users to have conversations with their favorite movie characters. It utilizes OpenAI's GPT API along with real movie dialogues stored in a database to generate realistic responses.

Features

• Level 1: Basic API Chatbot

  • A REST API (POST /chat) that accepts {character, user_message} and returns responses mimicking the character's personality.

• Level 2: Store & Retrieve Movie Script Data

  • Stores real movie dialogues in a database (MongoDB).
  • Retrieves exact or similar lines before generating responses.

• Level 3: Implement RAG with Vector Search

  • Converts movie scripts into vector embeddings using Pinecone.
  • Uses similarity search to fetch relevant dialogues before AI response generation.

• Level 4: Scale System to Handle High Traffic

  • Implements Redis caching for fast retrieval.
  • Adds rate limiting (5 requests/sec per user).
  • Uses async processing for performance optimization.
  • Load testing using K6.

• Level 5: Optimize for Latency & Deploy

  • Implements WebSockets for real-time chat.
  • Deploys the chatbot on Vercel.
  • Adds monitoring tools (Prometheus + Grafana).
  • Stores user chat history in a database.

Tech Stack

• Backend: Node.js, Express.js
• Database: MongoDB, Pinecone (Vector Database)
• Caching & Rate Limiting: Redis, express-rate-limit
• AI Model: OpenAI GPT API (or Llama2)
• WebSockets: Socket.io
• Monitoring: Prometheus, Grafana
• Load Testing: K6

Installation & Setup

Prerequisites

• Node.js & npm installed
• MongoDB database set up
• Pinecone API key (for vector search)
• Redis server running (for caching & rate limiting)
• OpenAI API key (if using GPT)

Steps

1. Clone the repository:

   git clone https://github.com/Waseb-untwale/AI-movie-character-chatbot
   cd movie-chatbot

2. Install dependencies:

   npm install

3. Set up environment variables in .env file:

   OPENAI_API_KEY=your_openai_api_key
   MONGO_URI=your_mongodb_connection_string
   PINECONE_API_KEY=your_pinecone_api_key

4. Start the application:

   npm start

API Endpoints

1. Chat with Character

• Endpoint: POST /chat
• Request Body:

  {
    "character": "Iron Man",
    "user_message": "Hey, what's up?"
  }

• Response:

  {
    "character": "Iron Man",
    "response": "Stay cool, kid. The world isn't gonna save itself."
  }

Deployment

The chatbot is deployed on Vercel for production use.

Monitoring & Performance Optimization

• Prometheus + Grafana: To monitor API performance.
• Rate Limiting: Prevents abuse (5 requests/sec per user).
• Redis Caching: Speeds up dialogue retrieval.
• Load Testing with K6: Ensures the system handles high traffic.

Contributing

Feel free to fork this repository and submit pull requests with improvements!