DataSage: Multi-agent Collaboration for Insight Discovery with External Knowledge Retrieval, Multi-role Debating, and Multi-path Reasoning

2025-11-18

AI Summary

Existing data insight agents suffer from insufficient domain knowledge integration, shallow analytical depth, and high code-generation error rates. To address these limitations, we propose a large language model–based multi-agent collaboration framework featuring three key innovations: (1) external knowledge retrieval to enhance domain contextual understanding; (2) a multi-role deliberation mechanism that simulates complementary analytical perspectives to strengthen reasoning; and (3) multi-path logical reasoning coupled with automated code validation to improve the accuracy of both generated code and insight conclusions. Experimental evaluation on the InsightBench benchmark demonstrates that our approach significantly outperforms existing state-of-the-art agents across all difficulty levels—particularly in insight accuracy and analytical depth. The framework establishes a scalable, robust paradigm for automated data insight generation.

Technology Category

• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Multiagent Systems: Modeling other Agents
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI

Abstract

In today's data-driven era, fully automated end-to-end data analytics, particularly insight discovery, is critical for discovering actionable insights that assist organizations in making effective decisions. With the rapid advancement of large language models (LLMs), LLM-driven agents have emerged as a promising paradigm for automating data analysis and insight discovery. However, existing data insight agents remain limited in several key aspects, often failing to deliver satisfactory results due to: (1) insufficient utilization of domain knowledge, (2) shallow analytical depth, and (3) error-prone code generation during insight generation. To address these issues, we propose DataSage, a novel multi-agent framework that incorporates three innovative features including external knowledge retrieval to enrich the analytical context, a multi-role debating mechanism to simulate diverse analytical perspectives and deepen analytical depth, and multi-path reasoning to improve the accuracy of the generated code and insights. Extensive experiments on InsightBench demonstrate that DataSage consistently outperforms existing data insight agents across all difficulty levels, offering an effective solution for automated data insight discovery.

Problem

• Automated data insight discovery lacks sufficient domain knowledge utilization
• Existing data analytics agents suffer from shallow analytical depth limitations
• Current insight generation approaches produce error-prone code and insights

Innovation

• External knowledge retrieval enriches analytical context
• Multi-role debating mechanism deepens analytical perspectives
• Multi-path reasoning improves code and insight accuracy