FinHelp

Landing Page
Statistics
Portfolio
Financial report analyzer
Stock news

What it does

FinHelp is an educational platform for new beginner investors. With the help of Natural Language Processing (NLP), the web app provides update-to-date stock news and stage-of-the-art analyzing tools so that new investors have a good place to start with their first-ever investment. Users can query for the latest stock news and prices, while also being able to get a summarization of company-specific financial reports. On the top of that, users will be suggested with the best investing strategy based on their current portfolio and industries of interests

Inspiration

The current inflation rate is 5% in the US while the wage increase is only at 3%, which means every US worker lose 2% of their total income every year on average. Therefore, investing in the stock market is a powerful tool eradicate that annual lost since stock market appreciate 8% of total return on average. We want FinHelp to give people more insights on the stocks market and possible investing strategy, so that newcomers to the field is not overweigh with tons of information about the stock market currently. We hope that FinHelp will help increase the number of Americans who own stocks (currently below 50%) to close to 100% so that nobody has to fight inflation

How we built it

Backend: Python Flask, Google Firebase
Frontend: Figma, React TypeScript
APIs: OpenAI, finBERT, Yahoo Finance APIs

Challenges we ran into

Limitation on current Natural Language Processing (NLP) model: the current model finBERT that we are utilizing is only able to tell if a sentence is positive or negative. And a financial report is more than 10 pages long. Therefore, we have to run the financial report through a summarization model offered by OpenAI, then use NLP on the summarized report.
Training our own model: At the beginning of the project, we are very ambitious on implement our own model to be able to analyze lengthy financial reports. it is definitely not an easy task for non-machine learning engineers.

Accomplishments that we're proud of

We're very proud of successfully integrating multiple artistic visions into one project. We were able to create an all-in-one hub for beginner investors where they can buy/sell and learn about certain stocks and investing strategies through the power of AI

What we learned

Each member of the team utilized challenging technology, and as a result, learned a lot about React and finBERT models during the last 6 days! We learned how to design, train and test hand finBERT models and build progressive web app in React.

This project really helped us have a better understanding of many of the capabilities within React, finBERT and in utilizing natual language processing to interpret company performance to use in an educational setting. We learned so much through this project and from each other and had a really great time working as a team!