RA-CFGPT: Chinese financial assistant with retrieval-augmented large language model

Jiangtong Li, Yang Lei, Yuxuan Bian, Dawei Cheng, Zhijun Ding, Changjun Jiang

Abstract: In this paper, we present the RA-CFGPT system, designed for Chinese financial tasks such as question answering, document analysis, investment advising, and risk assessment. Our system combines a hybrid knowledge base, a fine-tuned Chinese FinLLM, an information organizer, and response checkers to ensure outputs are accurate, compliant, and highlight potential risks. Experimental result reveals the effectiveness of our retrieval-augment training and RA-FGPT system.