AlphaStock AI ( KCG)

Inspiration We built AlphaStock AI because investing can be confusing with so much data and news. We wanted a simple way to help people understand the market and make smarter decisions using AI. What it does AlphaStock AI is a dashboard that shows real-time stock data, tracks your portfolio, and analyzes news. It includes an AI chatbot that can answer questions and a trading simulator that uses reinforcement learning to test strategies. How we built it We used Python with NiceGUI and Streamlit for the interface. We connected to Yahoo Finance for data, used scikit-learn for predictions, Stable-Baselines3 for reinforcement learning, and Gemini API for the chatbot. We also used SQLite to store user data. Challenges we ran into We had issues with API limits, installing dependencies, and making everything work together smoothly. It was also challenging to get the chatbot to answer both general and finance questions correctly. Accomplishments that we're proud of We built a full AI-powered financial platform that combines real-time data, machine learning, and a chatbot in one place. We also created a working RL trading simulator. What we learned We learned how to connect APIs, build AI features into apps, and apply machine learning and reinforcement learning to real problems. We also improved our debugging and development skills. What's next for AlphaStock AI (KCG) We want to improve the AI, make the app faster, add real trading integrations, and possibly create a mobile version.

Built With

• alpaca
• api
• fastapi
• feedparser
• gemini
• gymnasium
• matplotlib
• nicegui
• numpy
• pandas
• plaid
• plotly
• ppo
• python
• robinhood
• rss
• scikit-learn
• sqlite
• stable-baselines3
• streamlit
• yfinance