Earningly

Our project is centered around designing an innovative Natural Language Processing tailored to process and understand financial data. To achieve this, we're utilizing artificial intelligence technologies provided by Google, and harnessing rich data from the U.S. Securities and Exchange Commission (SEC). We've chosen Google's PaLM 2 as our main NLP model due to its state-of-the-art capabilities in language interpretation. NLP, an AI discipline, empowers computers to comprehend and interact with human language. With its understanding of language nuances, PaLM 2 can interpret and process financial documents, and it can then fetch and rephrase information to the user. At the heart of our project are SEC filings, including forms 10-K, 10-Q, and 8-K. To ensure our system always accesses the most recent data, we're tapping into the SEC's EDGAR system. EDGAR, or Electronic Data Gathering, Analysis, and Retrieval system, is a digital database provided by the SEC. This platform is where all U.S. publicly traded companies are mandated to submit their regulatory filings. These documents are essential sources of financial data. The 10-K form provides a comprehensive annual overview of a company's financial performance. The 10-Q form updates on the company's financial status on a quarterly basis, while the 8-K reports unscheduled material events or corporate changes that could be significant to shareholders or the SEC. Through processing these forms, our system can extract crucial financial details for each company, such as revenue, net income, and cash flow. This extraction is invaluable for investors, analysts, and financial institutions who need the most current, detailed financial information to make informed decisions. In addition, these filings are often hundreds of pages long, so an LLM that integrates directly into the source of the filing can efficiently fetch relevant information for our users.