Stock Analysis Dashboard

Description

In this project, we build a stock analysis dashboard. Key features include:

1. LLM-Powered Sentiment Analysis: Enter any stock symbol to retrieve recent news articles associated with it. Our integrated Large Language Model (LLM) analyzes these articles to provide insightful sentiment analysis, offering a deeper understanding beyond raw numbers.
2. TradingView Integration: We've embedded interactive widgets from TradingView to display real-time stock prices and general market information for the queried symbols, keeping you informed on current performance.

Deployment

Link to the deployed web app: Stock Sentiment Web App

Note: The web app is deployed on a Digital Ocean Droplet.

Docker Hub Images

• Web App: Link to Web App Docker Hub Image
• LLM Service: Link to LLM Service Docker Hub Image

Team Members

• Alan Chen
• Jackson Chen
• Ray Huang
• Ethan Yu

Project Structure

├── common/                  # Shared code between subsystems
│   └── models.py            # Database models 
├── llm/                     # LLM service for sentiment analysis
│   ├── Dockerfile           # Container configuration
│   ├── requirements.txt     # Python dependencies
│   ├── agent.py             # Core LLM agent logic
│   └── llm_app.py           # FastAPI application for the LLM service
│   └── tool.py              # Tools/functions used by the LLM agent
├── web-app/                 # Web application (user interface)
│   ├── Dockerfile           # Container configuration
│   ├── requirements.txt     # Python dependencies
│   ├── app.py               # FastAPI application for the web UI
│   └── templates/           # HTML templates for the web UI
│       ├── index.html
│       ├── detail.html
│       └── trending.html
├── docker-compose.yml       # Docker Compose configuration
├── LICENSE                  # Project License
├── pyproject.toml           # Python project configuration
└── README.md                

Setup and Configuration

1. Clone the Repository:

   git clone https://github.com/software-students-spring2025/5-final-galaxy.git
   cd 5-final-galaxy

2. Environment Variables:

   Create a .env file in the project root directory by copying the structure from .env.example . Fill in the required values. Based on the docker-compose.yml, you will need at least the following variables:

   # .env
   # Example MongoDB connection string
   MONGO_URI=mongodb://mongodb:27017/stockDB 
   # Choose your preferred LLM provider (Choose from OPENAI, GEMINI, XAI)              
   LLM_API_PROVIDER=OPENAI
   # Your API key for the LLM provider
   OPENAI_API_KEY=YOUR_OPENAI_API_KEY        
   GEMINI_API_KEY=YOUR_GEMINI_API_KEY        
   XAI_API_KEY=YOUR_XAI_API_KEY      

3. Docker: Ensure you have Docker and Docker Compose installed on your system.

Running the Project

1. Build and Run Containers: From the project root directory, run:

   docker-compose up --build -d

2. Accessing the Application:

   • The Web App should be accessible at http://localhost:5001.
   • The LLM service runs internally and is accessed by the Web App at http://llm:5002.
   • The MongoDB database is accessible on the host machine at mongodb://localhost:27017 if needed for direct inspection, but primarily used internally by the services.

3. Stopping the Application: To stop the running containers, execute:

   docker-compose down