# FinRobot: An Open-Source AI Agent Platform for Financial Applications using Large Language Models

**Authors:** Hongyang Yang, Boyu Zhang, Neng Wang, Cheng Guo, Xiaoli Zhang, Likun Lin, Junlin Wang, Tianyu Zhou, Mao Guan, Runjia Zhang, Christina Dan Wang  
**arXiv:** [2405.14767](https://arxiv.org/abs/2405.14767)  
**Date:** 2024-05-23  
**Summary page:** https://www.summarizepaper.com/en/paper/2405.14767v2/

## Key points

- Integration of Large Language Models (LLMs) in financial technology is increasing among institutions and professionals.
- Barriers like proprietary data and specialized knowledge hinder collaboration between the finance sector and the AI community.
- Researchers led by Hongyang Yang, Boyu Zhang, Neng Wang, Cheng Guo, Xiaoli Zhang, Likun Lin, Junlin Wang, Tianyu Zhou, Mao Guan, Runjia Zhang, and Christina Dan Wang are working on bridging this gap with FinRobot.
- FinRobot is an open-source AI agent platform tailored for financial applications using LLMs.
- The architecture of FinRobot includes four key layers: Financial AI Agents layer, Financial LLM Algorithms layer, LLMOps and DataOps layer, Multi-source LLM Foundation Models layer.
- FinRobot aims to democratize access to advanced AI tools for financial decision-making through open-source initiatives.
- It caters to both professional analysts and laypersons for utilizing powerful AI techniques in financial analysis.

## Summary

In the rapidly evolving landscape of financial technology, the integration of Large Language Models (LLMs) has become increasingly prevalent among financial institutions and professionals. However, significant barriers such as proprietary data and specialized knowledge continue to hinder collaboration between the finance sector and the AI community, limiting the potential for enhancing financial tasks effectively. Recognizing the pivotal role of financial analysis, a group of researchers led by Hongyang Yang, Boyu Zhang, Neng Wang, Cheng Guo, Xiaoli Zhang, Likun Lin, Junlin Wang, Tianyu Zhou, Mao Guan, Runjia Zhang, and Christina Dan Wang have embarked on a mission to bridge this gap. Their groundbreaking initiative involves the development of FinRobot: an innovative open-source AI agent platform tailored specifically for financial applications utilizing LLMs. This platform is designed to support multiple financially specialized AI agents powered by LLM technology. The architecture of FinRobot comprises four key layers1) Financial AI Agents layer: This layer focuses on formulating Financial Chain-of-Thought (CoT) by breaking down complex financial problems into logical sequences. 2) Financial LLM Algorithms layer: Dynamically configuring appropriate model application strategies for specific tasks. 3) LLMOps and DataOps layer: Producing accurate models through the application of training/fine-tuning techniques and leveraging task-relevant data. 4) Multi-source LLM Foundation Models layer: Integrating various LLMs to enable seamless access for the layers above. By democratizing access to these advanced AI tools through open-source initiatives like FinRobot (available at https://github.com/AI4Finance-Foundation/FinRobot), the team aims to promote wider adoption of AI in financial decision-making processes. Notably, FinRobot caters to both professional-grade analysts and laypersons alike, providing hands-on experience in utilizing powerful AI techniques for sophisticated financial analysis. This whitepaper serves as a comprehensive guide to understanding FinRobot's capabilities and its potential impact on revolutionizing financial technology. With a diverse team of experts at its helm and a commitment to open collaboration within the industry, FinRobot represents a significant step towards unlocking the full potential of AI in finance.