CogniCents

Inspiration

The inspiration for CogniCents came from the realization that the average person struggles with managing their finances effectively. We wanted to create an AI-powered finance app that could provide personalized financial guidance and make it easier for people to take control of their finances.

What it does

CogniCents currently has the ability to provide the user with summarization into layman's terms from technical finance jargon. Additionally, it can take in the current 10-Q reports from each day to perform a similar summative analysis. CogniVest can also help the average to hardcore investors stay up-to-date on the latest financial news and trends. It will provide the users with real-time data visualization - providing easy-to-understand graphical representations of market trends for any listed company the user wishes to provide. This allows users to quickly identify patterns and make informed investment decisions. Real-time finance articles from Fidelity API - Keeps users up-to-date with the latest news and events from a reliable source, providing valuable insights into market trends and potential investment opportunities. News Sentiment Analysis scoring is included to provide users with a score that helps them assess market trends and adjust their investment strategies accordingly. This feature is powered by AI and NLP technology, allowing users to access sophisticated analysis tools that can help them make more informed decisions.

How we built it

We built the CogniCents user interface with NextJS as a front-end framework. We used Python with FastAPI in the back end to facilitate using machine learning libraries. We utilized various HuggingFace models including one trained using Amazon SageMaker and the new Hugging Face Deep Learning container and the DistilBERT model, a Smaller, faster, cheaper, lighter, adistilled version of BERT. We also utilized calls to OpenAI's gpt-3.5-turbo.

Challenges we ran into

At the beginning of the hackathon, we ran into very many issues with running boilerplate code on each person's machine. This ended up using up a good chunk of our time at the hackathon.

Accomplishments that we're proud of

We are proud of developing an AI-powered finance app despite all of our teammates meeting for the first time at the hackathon, as well as the countless hours of issues that each of our computers had at the beginning of the event.