Stock Market AI Agent / Dashboard

A modern, AI-powered stock market dashboard and chat agent for real-time and historical analysis of S&P 500 stocks, built with Streamlit, LangGraph & Yahoo Finance.

Features

Real-time and historical stock data fetching using YFinance
AI-powered stock analysis, chat, and recommendations
Interactive Streamlit dashboards for S&P 500 and individual stocks
Automated local data storage and updates
Technical indicators, financial ratios, and analyst recommendations
Latest news and sentiment integration

Project Structure

app.py — Streamlit app entry point and navigation.
pages/ — Chat agent, S&P 500 monitor, and symbol monitor dashboards.
- Chat Agent: Uses deepseek-r1-distill-llama-70b LLM, equipped with a few tools around Yahoo Finance API provided by Agno as well as some custom tools.
- S&P 500 Monitor: Uses automated logic to rank stocks with most significant changes in stock prices for a specified period. Logic flow:
  - Loading data from local files.
  - Calculating various changes such as 1h, 6h, 12h, 2d.
  - Ranking top n stock based on based on highest changes.
- Symbol Monitor: Can be used to explore trends in a single stock.
data_fetcher.py — Automated data download and update logic that keeps running in background. Currently, it only fetches last 5 days of data, with 1 minute of interval during open-market times.
tools.py — LangChain tools for stock analysis, news, and calculating changes in stock prices.
S&P.json — List of S&P 500 companies and tickers.
data/ — Local storage for stock data (organized by symbol).

Prerequisites

Python 3.10
API keys:
- Groq for LLM inference.
- Serper for Google search tool.

Installation

Clone this repository:

git clone https://github.com/byahmedali/stockmarketagent.git
cd stockmarketagent

Create a .env file

GROQ_API_KEY=YOUR_GROQ_API_KEY
SERPER_API_KEY=YOUR_SERPER_API_KEY

Set up a virtual environment

a. Create a new virtual environment:
```
py -3.10 -m venv venv
```
b. Activate the virtual environment:

Linux/macOS:
```
source venv/bin/activate
```
Windows:
```
venv\Scripts\activate.bat
```
c. Confirm you're using the newly created environment:

If you see environment name (venv) at the beginning of your command prompt, it confirms that you're using the newly created virtual environment.
Install required dependencies:
```
pip install -r requirements.txt
```

Usage

Run data_fetcher.py in background:
```
python data_fetcher.py
```
Note: On first launch, data fetcher will take some time.
Open a new terminal in the same environment and start the Streamlit app:
```
streamlit run app.py
```
Open your browser and go to http://localhost:8501

Navigate between the Chat Agent, S&P 500 Monitor, and Symbol Monitor from the sidebar.

Running with Docker

Build the Docker image:
```
docker build -t stockagent .
```
Run the container:
```
docker run -p 8501:8501 --env-file .env stockagent
```
Note: It is essential to provide API keys when running the docker container.
Open your browser and go to http://localhost:8501

Contributing

New idea and feature contributions are always welcome!

Acknowledgments

This work is a part of ATS's projects catalog.
Yahoo Finance & Agno contributors
Python & LangChain community

byahmedali/stockmarketagent