StockSim

Inspiration

A barrier that keeps people from investing is the fear of losing money. Most beginners want to learn how the stock market works, but the idea of losing real money the first time they try is enough to stop them before they start. Traditional apps like Robinhood offer a great platform but no guidance. They don't teach people how to capture how the markets actually moves. We wanted to build something that felt like a real market, with real volatility and unpredictability, but with the safety net fully intact.

What it does

StockSim is a simulation platform where users can buy and sell stocks in a completely risk-free environment with their own finances and spending habits built in. We use both historical replays and also generate market conditions using Gen-AI that simulates real-world trading behavior, price movement, and market trends. Users pick stocks, run simulations over a chosen time period, and watch how their decisions would affect their net stock worth. All your data stays local and safe on your computer.

How we built it

Frontend: HTML/CSS, Javascript Backend: Flask Python, Yfinance, Gemini, Nessie, MCP Database: TigerData PostgreSQL

Challenges we ran into

Getting a smooth motion into the charts which involved smart caching and loading to accomplish.
Learning about trading strategies so we can make an effective game.
Incorporating and finding news and accurately simulating multiple environments

Accomplishments that we're proud of

The implementation of multiple trading strategies with accurate historical data to simulate real markets.
The helpful tips given by the AI agent deepening the understanding of the strategies.
The simulation of new data by the gemini api to simulate crashes and other events that you have to accurately predict and survive.
The use of nessie to bring a personal aspect by using "your" personal details and spending habits to gamify and personalize the learning experience.

What we learned

Trading Strategies
Market volatility and analysis
Algorithmic Trading
Gen-AI use in teaching multi-faceted concepts
Model Context Protocol and how LLMS interact with them.

What's next for StockSim

Serving as a Stock Trading platform using Yfinance web sockets to trade real stocks after they have been trained with our platform.
Adding more strategies that people can learn.
Using the MCP server and the simulated stock market to see how trading models might perform in the market.