🤖 Gemini Multi-Modal Chat Application

A sophisticated multi-modal chat application built with Streamlit and FastAPI, powered by Google's Gemini AI model - gemini-2.5-flash. This application supports intelligent conversations across text, images, and CSV data analysis — all in one seamless interface.

✨ Features

💬 Core Chat Mode

• Multi-turn conversational chat with persistent memory
• Real-time streaming responses for natural conversation flow
• Markdown rendering support for rich text formatting
• Session-based chat history management

🖼️ Image Chat Mode

• Upload and analyze images using Gemini's vision capabilities
• Ask natural language questions about image content
• Supports .jpg, .jpeg, .png formats
• Secure base64 encoding for backend transfer

📊 CSV Chat Mode

• Upload CSV files or import from URLs
• Intelligent data analysis with natural language queries
• Automatic data preview and summary statistics
• Interactive plotting and visualization support
• Context-aware data exploration powered by Gemini AI

🚀 Quick Start

📋 Prerequisites

• Python 3.9.8
• A valid Google Gemini API key (Get one here)
• Supabase to store chat histories.

⚙️ Installation

1️⃣ Clone the repository

git clone https://github.com/C-zu/gemini-chatapp.git
cd Chatbot

2️⃣ Create a virtual environment

python -m venv .venv

# Activate it
# macOS / Linux
source .venv/bin/activate
# Windows
.venv\Scripts\activate

3️⃣ Install dependencies

pip install -r requirements.txt

or

pip install fastapi langchain google-generativeai plotly langchain_google_genai streamlit dotenv supabase langchain_experimental uvicorn tabulate

4️⃣ Set up environment variables

Create a .env file in the project root and add your Gemini API key:

GOOGLE_API_KEY=your_gemini_api_key_here
SUPABASE_URL=your_supabase_url_here
SUPABASE_KEY=your_supabase_key_here

▶️ Running the Application

🎨 Frontend (Streamlit)

cd frontend
streamlit main.py

⚡ Backend (FastAPI)

cd backend
py main.py

Once both services are running:

• Frontend: Open your browser at http://localhost:8501
• Backend API: Available at http://localhost:8000
• API Documentation: Visit http://localhost:8000/docs for interactive API docs

📁 Project Structure

Chatbot/
├── frontend/                    # Streamlit frontend application
│   ├── main.py                 # Main Streamlit app entry point
│   ├── core_chat.py            # Core text chat functionality
│   ├── image_chat.py           # Image analysis chat mode
│   ├── csv_chat.py             # CSV data analysis mode
│   ├── components/
│   │   └── sidebar.py          # Sidebar navigation component
│   ├── services/
│   │   └── api_client.py       # FastAPI backend communication
│   └── utils/
│       ├── session_manager.py  # Chat session handling
│       ├── database_setup.py   # Database initialization
│       └── helpers.py          # Utility functions
│   
│   
│
├── backend/                     # FastAPI backend application
│   ├── main.py                 # FastAPI app entry point
│   ├── api/
│   │   ├── ai.py               # AI chat endpoints
│   │   ├── chat.py             # Chat management endpoints
│   │   └── sessions.py         # Session management endpoints
│   ├── services/
│   │   ├── chat_service.py     # Chat business logic
│   │   ├── chat_repository.py  # Chat data persistence
│   │   └── session_service.py  # Session management logic
│   ├── models/
│   │   └── schemas.py          # Pydantic data schemas
│   └── utils/
│       └── supabase_client.py  # Database client configuration
│
├── data/                        # Data storage directory
│   └── chats/                  # Chat session data files
│
├── requirements.txt            # Python dependencies
├── chainlit.md                 # Chainlit welcome screen
└── README.md                   # This file

🔌 API Endpoints

Endpoint	Method	Description
/api/v1/ai/chat	POST	Stream text-based chat responses
/api/v1/ai/chat/image	POST	Stream image analysis responses
/api/v1/ai/chat/csv	POST	CSV data analysis with plotting support
/api/v1/sessions/{session_id}/messages	GET	Retrieve messages from a session
/api/v1/sessions/{session_id}/messages	POST	Add a new message to a session
/health	GET	Health check endpoint

🏗️ Architecture

Frontend (Streamlit)

• UI Rendering: Clean, responsive interface with mode switching
• File Handling: Upload and processing of images and CSV files
• Session Management: Persistent chat sessions with history
• Real-time Communication: Streaming responses from backend

Backend (FastAPI)

• API Gateway: RESTful endpoints with automatic documentation
• AI Integration: Seamless integration with Google Gemini models
• Data Processing: CSV analysis with pandas and plotly visualization
• CORS Support: Cross-origin requests for frontend communication

AI Service (Gemini)

• Text Processing: Natural language understanding and generation
• Vision Analysis: Image content analysis and description
• Data Analysis: Intelligent CSV data exploration and insights
• Streaming Responses: Real-time response generation

🛠️ Technology Stack

• Frontend: Streamlit, Python
• Backend: FastAPI, Uvicorn
• AI/ML: Google Gemini API, LangChain
• Data Processing: Pandas, Plotly
• Database: Supabase

🔧 Configuration

Environment Variables

• GOOGLE_API_KEY: Your Google Gemini API key
• SUPABASE_URL: Supabase project URL
• SUPABASE_KEY: Supabase API key

Model Configuration

The application uses gemini-2.5-flash by default. You can modify the model in backend/services/ai_service.py:

def __init__(self, model_name="gemini-2.5-flash", temperature=0.3):

Demo video link:

https://www.youtube.com/watch?v=ZVQd3i5MBRI