RBPA: Retrieval-Augmented-Generation based Personal Investment Assistant
Keywords: Investment assistant, RAG, Reasoning
Abstract: RBPA employs Retrieval-Augmented Generation (RAG) to provide a sophisticated personal investment assistant. It addresses the limitations of Large Language Models (LLMs) in financial markets by integrating a knowledge base with real-time data retrieval. The system is designed to offer tailored investment advice by combining professional financial insights with individual investor data. RBPA's approach includes a graph database for comprehensive document analysis and methods to enhance LLM's logical capabilities in finance. It aims to deliver personalized and informed investment strategies to users. RBPA also involves a new metric and benchmark to evaluate the performance of LLM investment decision more comprehensively, includes both ROI and reasoning score.
Submission Number: 16
Loading