Various industries such as finance, meteorology, and energy generate vast amounts of heterogeneous data every day. There is a natural demand for humans to manage, process, and display data efficiently. However, it necessitates labor-intensive efforts and a high level of expertise for these data-related tasks. Considering that large language models (LLMs) have showcased promising capabilities in semantic understanding and reasoning, we advocate that the deployment of LLMs could autonomously manage and process massive amounts of data while displaying and interacting in a human-friendly manner. Based on this belief, we propose Data-Copilot, an LLM-based system that connects numerous data sources on one end and caters to diverse human demands on the other end. Acting like an experienced expert, Data-Copilot autonomously transforms raw data into visualization results that best match the user's intent. Specifically, Data-Copilot autonomously designs versatile interfaces (tools) for data management, processing, prediction, and visualization. In real-time response, it automatically deploys a concise workflow by invoking corresponding interfaces step by step for the user's request. The interface design and deployment processes are fully controlled by Data-Copilot itself, without human assistance. Besides, we create a Data-Copilot demo that links abundant data from different domains (stock, fund, company, economics, and live news) and accurately respond to diverse requests, serving as a reliable AI assistant.

翻译：摘要：金融、气象和能源等各行各业每天都会产生大量异构数据，人类对这些数据进行高效管理、处理和展示的需求自然存在。然而，完成这些数据相关任务需要耗费大量人力且要求较高的专业知识。鉴于大型语言模型（LLMs）在语义理解和推理方面展现出令人瞩目的能力，我们认为部署LLMs可以自主管理和处理海量数据，同时以人类友好的方式进行展示和交互。基于这一理念，我们提出Data-Copilot，一个基于LLM的系统，一端连接众多数据源，另一端满足多样的人类需求。Data-Copilot像经验丰富的专家一样，自动将原始数据转化为最符合用户意图的可视化结果。具体而言，Data-Copilot自主设计用于数据管理、处理、预测和可视化的多功能接口（工具）。在实时响应中，它通过逐步调用相应接口，自动为用户请求部署简洁的工作流。接口的设计和部署过程完全由Data-Copilot自主控制，无需人工协助。此外，我们创建了Data-Copilot演示系统，该系统连接来自不同领域（股票、基金、公司、经济及时事新闻）的丰富数据，并能准确响应各种请求，作为可靠的AI助手。