Project Apollo

Landing page with a search bar that is capable of auto-filling queries
Generative AI report about the company's performance and our model's prediction
Visual representation of the company's performance data

Inspiration

Guided by the Greek God of prophecy, Apollo, we aim to see into the future and help consumers become aware about the companies. Our mission is to connect consumers to companies in the simplest way possible.

What it does

Our app analyzes companies' financial performance through revenue reports, news headlines, and time series forecasting, to predict the company's performance for the upcoming financial quarter.

How we built it

We used the Yahoo Finance API to collect revenue reports for the company, MarketAUX to collect snippets of new relevant to the company, and a python script that utilizes time series forecasting to predict performance. The data from these 3 sources are then fed into our Llama 2 Generative AI model that analyses all the information and outputs a detailed report about predicted performance of the company.

Challenges we ran into

Some of the main challenges we ran into were:

Dealing with cost and space issue for LLM
Finding good data for stock analysis
Creating accurate prompts for our model

Accomplishments that we're proud of

We were able to successfully build a working ensemble model, which we are very proud of. Additionally, we also created the dynamic react app, which was rather tedious but we were able to finish before the deadline.

What we learned

We learned about the importance of deployment and gained lots of insightful knowledge about sentiment analysis, working with Large Language Models (LLMs), and integrating generative AI in our application.