Research paper proposes RAG-based LLM system for automated investor briefs using GPT-4o

The paper, titled 'Augmenting Fundamental Analysis with Large Language Models: A RAG-Based System for Generating Investor Briefs,' explores using LLMs to automate aspects of fundamental analysis. It processes company reports, macroeconomic indicators (GDP, inflation), and SEC filings from EDGAR. The data is preprocessed and sent to OpenAI's GPT-4o model in a RAG-like regime. The authors also created a document describing an exemplar investor knowledge base based on Kitchin cycles. The system aims to generate concise investor briefs, potentially reducing manual analysis time.