DRFLOW: A Deep Research Benchmark for Personalized Workflow Prediction

2026-06-16

AI Summary

This work addresses the limitations of existing deep research systems, which primarily focus on report generation and summarization yet fail to meet enterprise demands for personalized, actionable workflow prediction. We propose DRFLOW—the first benchmark tailored for personalized workflow forecasting—encompassing 100 cross-domain tasks, 1,246 annotated steps, and over 3,900 heterogeneous information sources. To rigorously evaluate performance, we introduce a seven-dimensional metric assessing factual grounding, step reconstruction, structural ordering, conditional reasoning, and personalization, among other aspects. Built upon a retrieval-augmented architecture, our DRFLOW-Agent integrates multi-source evidence extraction, step-sequence modeling, and personalized inference. Experimental results demonstrate that the agent outperforms strong baselines by up to 10.02% in average F1 score, thereby validating both the challenge posed by the task and the effectiveness of the proposed benchmark.

Abstract

Deep research (DR) systems are increasingly used for complex information-seeking tasks, but existing works mainly focus on generating reports and summaries. In contrast, many enterprise tasks instead require an agent to identify concrete workflows which is a sequence of action-steps. For example, rather than summarizing budgeting policies, an agent should be able to determine the steps needed to answer a question such as: "How do I request new headcount given a fixed budget?". Therefore, we introduce DRFLOW, a benchmark for evaluating personalized workflows predicted by agents from heterogeneous sources. Each task requires the agent to identify relevant evidence from scattered sources, then use that evidence to predict the correct action-step sequence for the user's task. DRFLOW contains 100 tasks across five domains, with 1,246 reference workflow steps grounded in more than 3,900 sources. We define seven diagnostic metrics covering factual grounding, step recovery, structural ordering, condition resolution, and personalization. We further present DRFLOW-Agent (DRFA), a workflow-oriented reference agent to predict personalized workflow. We show that although DRFA improves over strong baseline agents (upto 10.02% average F1 score), there is substantial room for improvement remains across these workflow metrics, indicating that predicting complete and correct personalized workflows remains a challenging frontier for deep research.

Problem

• personalized workflow prediction
• deep research
• action-step sequence
• heterogeneous sources
• workflow grounding

Innovation

• personalized workflow prediction
• deep research benchmark
• action-step sequence
• heterogeneous information sources
• workflow grounding