EduStock

Helping Young Adults Break into The Market. We use an AI assisted financial assistant who gives financial advice on the stock market. We also incorporate advance AI models to predict the market.

Comment

Inspiration

We started this journey with a shared dream: to break into the world of quantitative finance. The markets have always felt like a grand puzzle, complex, fast-moving, and endlessly evolving. What inspired us was not just the idea of building a project, but of building a stepping stone toward becoming the kind of quants who could solve that puzzle.

What it does

EduStock is a research tool that surfaces trade ideas. It takes in recent news headlines and price data, scores sentiment, extracts topological features from time-series (via persistent homology), and turns it into a ranking of tickers with buy/sell lean and brief rationales. It’s decision support, not financial advice.

How we built it

We built a pipeline that collects headline and OHLCV data, computes sentiment from titles, derives topological descriptors of price windows (persistence summaries), and trains a lightweight model to map those features to next-day detection.

Challenges we ran into

Gathering headline data was hard due to the web scraping rate limits. Also, the initial learning curve of using topological analysis and the math behind it was challenging but fun!

Accomplishments that we're proud of

  • We built and trained our own next-day returns prediction model
  • We web scraped 4 million articles efficiently by distributing our workload over the Temple computing network infrastructure ## What we learned We learned how to web scrape for data, using higher level math and data analysis, and teamwork ## What's next for EduStock Expand coverage to crypto, full articles rather than just title

Built With

Share this project:

Updates