🤖 QA System using Gemini Pro API

A powerful Question and Answer system built with Google's Gemini Pro API, featuring vector storage with AstraDB and LLM monitoring with LangSmith.

✨ Features

• 💬 Text Q&A: Ask questions and get detailed answers powered by Gemini Pro
• 🖼️ Image Analysis: Upload images for AI-powered analysis with Gemini Pro Vision
• 📄 Document Processing: Extract and analyze text from PDF, DOCX, and other file formats
• 🔗 URL Processing: Analyze content from web pages and YouTube videos
• 🔍 Semantic Search: Find similar content using vector embeddings
• 📊 LLM Monitoring: Track and analyze model performance with LangSmith
• 🗄️ Vector Storage: Store and retrieve vectors using AstraDB
• 🌐 Web Interface: Clean, responsive Flask web application

🚀 Tech Stack

Technology	Purpose
	Core language
	Web framework
	LLM model
	Vector database
	LLM monitoring
	Text embeddings
	Containerization
	Cloud hosting

🏗️ Architecture

┌─────────────────┐     ┌─────────────────┐     ┌─────────────────┐
│                 │     │                 │     │                 │
│  Web Interface  │────▶│  Flask Backend  │────▶│   Gemini Pro    │
│                 │     │                 │     │                 │
└─────────────────┘     └────────┬────────┘     └─────────────────┘
                               │
                               ▼
                 ┌─────────────────────────────┐
                 │                             │
                 │  File/URL/Image Processing  │
                 │                             │
                 └──────────────┬──────────────┘
                               │
                 ┌─────────────▼──────────────┐
                 │                            │
                 │  Sentence Transformers     │
                 │  (Vector Embeddings)       │
                 │                            │
                 └──────────────┬─────────────┘
                               │
          ┌──────────────────┐ │ ┌───────────────────┐
          │                  │ │ │                   │
          │     AstraDB      │◀┴─▶│    LangSmith      │
          │  (Vector Store)  │   │    (Monitoring)   │
          │                  │   │                   │
          └──────────────────┘   └───────────────────┘

🔧 Installation

Prerequisites

• Python 3.10 or higher
• Google API key for Gemini Pro
• AstraDB account and token
• (Optional) LangSmith API key

Option 1: Using Setup Scripts (Recommended) 🚀

1. Clone the repository:

   git clone https://github.com/VisionExpo/QA-System-using-Gemini-Pro-API.git
   cd QA-System-using-Gemini-Pro-API

2. Run the setup script:

   For Windows:

   setup_env.bat

   For macOS/Linux:

   chmod +x setup_env.sh
   ./setup_env.sh

   This script will:

   • 🔨 Create a virtual environment
   • ⚡ Activate the virtual environment
   • 📦 Install dependencies
   • 🔑 Create a .env file from the example if it doesn't exist

3. Edit the .env file and add your API keys:

   • GOOGLE_API_KEY: Your Google API key
   • ASTRA_DB_TOKEN: Your AstraDB token
   • ASTRA_DB_ENDPOINT: Your AstraDB endpoint
   • LANGCHAIN_API_KEY: (Optional) Your LangSmith API key

Option 2: Manual Setup 🛠️

1. Clone the repository:

   git clone https://github.com/VisionExpo/QA-System-using-Gemini-Pro-API.git
   cd QA-System-using-Gemini-Pro-API

2. Create and activate a virtual environment:

   For Windows:

   python -m venv venv
   venv\Scripts\activate

   For macOS/Linux:

   python -m venv venv
   source venv/bin/activate

3. Install dependencies:

   pip install -r requirements.txt

4. Set up your environment variables:

   cp .env.example .env

   Then edit the .env file to add your API keys.

🚀 Usage

Running the Application

python app.py

Then open http://127.0.0.1:5000 in your web browser.

💬 Text Q&A Mode

Ask any question and get detailed answers from Gemini Pro:

• General knowledge questions
• Coding help
• Explanations of complex topics
• Creative writing assistance

🖼️ Image Analysis Mode

Upload images for AI-powered analysis:

• Object identification
• Scene description
• Text extraction from images
• Visual content analysis

📄 Document Processing

Upload and analyze documents:

• PDF files
• Word documents (DOCX)
• Text files
• CSV data

🔗 URL and YouTube Processing

Analyze content from:

• Web pages
• YouTube videos (transcripts and summaries)
• Online articles

🌐 Deployment

This application is deployed on Render. You can access it at: https://qa-system-using-gemini-pro-api-1.onrender.com/

For detailed deployment instructions, see RENDER_DEPLOYMENT.md.

🧪 Testing

To run tests:

python -m pytest tests/

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgements

• Google for the Gemini Pro API
• DataStax for AstraDB
• LangChain for LangSmith
• The open-source community for various libraries used in this project

📞 Contact

For questions or feedback, please open an issue on GitHub or contact the maintainer at gorulevishal984@gmail.com.

---

Made with ❤️ by Vishal Gorule