Enterprise Deep Research: Steerable Multi-Agent Deep Research for Enterprise Analytics
Akshara Prabhakar, Roshan Ram, Zixiang Chen, Silvio Savarese, Frank Wang, Caiming Xiong, Huan Wang, Weiran Yao
2025-10-21
Summary
This paper introduces a new system called Enterprise Deep Research, or EDR, which is designed to automatically gather and analyze information for businesses dealing with huge amounts of unstructured data.
What's the problem?
Companies are drowning in data – things like reports, articles, and code – but it’s hard to turn that raw information into useful insights. Current AI systems, especially those using 'agents' to do research, often struggle to understand the specific needs of a business, accurately interpret what’s being asked, and fit into existing company workflows. They also don't always know when they're missing key information.
What's the solution?
The researchers built EDR as a team of specialized AI agents working together. There's a 'master planner' that breaks down complex questions, then different agents focus on searching specific sources like the general web, academic papers, code repositories like GitHub, and professional networking sites like LinkedIn. EDR also includes tools to convert natural language into database queries, analyze files, and connect to company systems. A 'visualization agent' then presents the findings in a clear way, and the system can even recognize when it needs more information and ask for guidance, or adjust its search strategy. It can create reports automatically and provide real-time updates.
Why it matters?
This work is important because it shows how AI can be used to automate complex research tasks for businesses, leading to faster and more informed decision-making. EDR performs better than other similar systems, even without human help, and the researchers are sharing their code and data so others can build on their work and improve AI-powered research tools.
Abstract
As information grows exponentially, enterprises face increasing pressure to transform unstructured data into coherent, actionable insights. While autonomous agents show promise, they often struggle with domain-specific nuances, intent alignment, and enterprise integration. We present Enterprise Deep Research (EDR), a multi-agent system that integrates (1) a Master Planning Agent for adaptive query decomposition, (2) four specialized search agents (General, Academic, GitHub, LinkedIn), (3) an extensible MCP-based tool ecosystem supporting NL2SQL, file analysis, and enterprise workflows, (4) a Visualization Agent for data-driven insights, and (5) a reflection mechanism that detects knowledge gaps and updates research direction with optional human-in-the-loop steering guidance. These components enable automated report generation, real-time streaming, and seamless enterprise deployment, as validated on internal datasets. On open-ended benchmarks including DeepResearch Bench and DeepConsult, EDR outperforms state-of-the-art agentic systems without any human steering. We release the EDR framework and benchmark trajectories to advance research on multi-agent reasoning applications. Code at https://github.com/SalesforceAIResearch/enterprise-deep-research and Dataset at https://huggingface.co/datasets/Salesforce/EDR-200