Scaling Generalist Data-Analytic Agents

• URL: http://arxiv.org/abs/2509.25084v1
• Date: Mon, 29 Sep 2025 17:23:08 GMT
• Title: Scaling Generalist Data-Analytic Agents
• Authors: Shuofei Qiao, Yanqiu Zhao, Zhisong Qiu, Xiaobin Wang, Jintian Zhang, Zhao Bin, Ningyu Zhang, Yong Jiang, Pengjun Xie, Fei Huang, Huajun Chen,
• Abstract summary: DataMind is a scalable data synthesis and agent training recipe designed to build generalist data-analytic agents.<n>DataMind tackles three key challenges in building open-source data-analytic agents.
• Score: 95.05161133349242
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Data-analytic agents are emerging as a key catalyst for automated scientific discovery and for the vision of Innovating AI. Current approaches, however, rely heavily on prompt engineering over proprietary models, while open-source models struggle to face diverse-format, large-scale data files and long-horizon, multi-step reasoning that real-world analytics demands. This paper introduces DataMind, a scalable data synthesis and agent training recipe designed to build generalist data-analytic agents. DataMind tackles three key challenges in building open-source data-analytic agents, including insufficient data resources, improper training strategy, and unstable code-based multi-turn rollout. Concretely, DataMind applies 1) a fine-grained task taxonomy and a recursive easy-to-hard task composition mechanism to increase the diversity and difficulty of synthesized queries; 2) a knowledge-augmented trajectory sampling strategy followed by model-based and rule-based filtering; 3) a dynamically adjustable training objective combining both SFT and RL losses; 4) a memory-frugal and stable code-based multi-turn rollout framework. Built on DataMind, we curate DataMind-12K, a high-quality trajectory set spanning diverse domains, task categories, and data file formats for data-analytic tasks. Trained on DataMind-12K, our DataMind-14B achieves state-of-the-art with an average score of 71.16% on multiple data analysis benchmarks, outperforming the strongest proprietary baselines DeepSeek-V3.1 and GPT-5. Our DataMind-7B also performs best among all open-source models with a score of 68.10%. We also incorporate some empirical insights gained from our exploratory trials into the analysis experiments, aiming to provide actionable insights about agentic training for the community. We will release DataMind-12K and DataMind-7B,14B for the community's future research.

Related papers

• Can LLMs Clean Up Your Mess? A Survey of Application-Ready Data Preparation with LLMs [66.63911043019294]
  Data preparation aims to denoise raw datasets, uncover cross-dataset relationships, and extract valuable insights from them.<n>This paper focuses on the use of LLM techniques to prepare data for diverse downstream tasks.<n>We introduce a task-centric taxonomy that organizes the field into three major tasks: data cleaning, standardization, error processing, imputation, data integration, and data enrichment.
  arXiv  Detail & Related papers  (2026-01-22T12:02:45Z)
• OpenDataArena: A Fair and Open Arena for Benchmarking Post-Training Dataset Value [74.80873109856563]
  OpenDataArena (ODA) is a holistic and open platform designed to benchmark the intrinsic value of post-training data.<n>ODA establishes a comprehensive ecosystem comprising four key pillars: (i) a unified training-evaluation pipeline that ensures fair, open comparisons across diverse models; (ii) a multi-dimensional scoring framework that profiles data quality along tens of distinct axes; and (iii) an interactive data lineage explorer to visualize dataset genealogy and dissect component sources.
  arXiv  Detail & Related papers  (2025-12-16T03:33:24Z)
• DeepAnalyze: Agentic Large Language Models for Autonomous Data Science [35.69385623867138]
  We introduce DeepAnalyze-8B, the first agentic designed for autonomous data science.<n>We propose a curriculum-based agentic training paradigm that emulates the learning trajectory of human data scientists.<n>We also introduce a data-grounded trajectory synthesis framework that constructs high-quality training data.
  arXiv  Detail & Related papers  (2025-10-19T15:13:42Z)
• Multimodal Data Storage and Retrieval for Embodied AI: A Survey [8.079598907674903]
  Embodied AI (EAI) agents interact with the physical world, generating vast, heterogeneous multimodal data streams.<n>EAI's core requirements include physical grounding, low-latency access, and dynamic scalability.<n>Our survey is based on a comprehensive review of more than 180 related studies, providing a rigorous roadmap for designing the robust, high-performance data management frameworks.
  arXiv  Detail & Related papers  (2025-08-19T15:04:02Z)
• Why Do Open-Source LLMs Struggle with Data Analysis? A Systematic Empirical Study [55.09905978813599]
  Large Language Models (LLMs) hold promise in automating data analysis tasks.<n>Yet open-source models face significant limitations in these kinds of reasoning-intensive scenarios.<n>In this work, we investigate strategies to enhance the data analysis capabilities of open-source LLMs.
  arXiv  Detail & Related papers  (2025-06-24T17:04:23Z)
• Making Sense of Data in the Wild: Data Analysis Automation at Scale [0.1747623282473278]
  We propose a novel approach that combines intelligent agents with retrieval augmented generation to automate data analysis, dataset curation and indexing at scale.<n>We demonstrate that our approach results in more detailed dataset descriptions, higher hit rates and greater diversity in dataset retrieval tasks.
  arXiv  Detail & Related papers  (2025-01-27T10:04:10Z)
• Putting Data at the Centre of Offline Multi-Agent Reinforcement Learning [3.623224034411137]
  offline multi-agent reinforcement learning (MARL) is an exciting direction of research that uses static datasets to find optimal control policies for multi-agent systems. Though the field is by definition data-driven, efforts have thus far neglected data in their drive to achieve state-of-the-art results. We show how the majority of works generate their own datasets without consistent methodology and provide sparse information about the characteristics of these datasets.
  arXiv  Detail & Related papers  (2024-09-18T14:13:24Z)
• DSBench: How Far Are Data Science Agents from Becoming Data Science Experts? [58.330879414174476]
  We introduce DSBench, a benchmark designed to evaluate data science agents with realistic tasks.<n>This benchmark includes 466 data analysis tasks and 74 data modeling tasks, sourced from Eloquence and Kaggle competitions.<n>Our evaluation of state-of-the-art LLMs, LVLMs, and agents shows that they struggle with most tasks, with the best agent solving only 34.12% of data analysis tasks and achieving a 34.74% Relative Performance Gap (RPG)
  arXiv  Detail & Related papers  (2024-09-12T02:08:00Z)
• CMDBench: A Benchmark for Coarse-to-fine Multimodal Data Discovery in Compound AI Systems [10.71630696651595]
  Compound AI systems (CASs) that employ LLMs as agents to accomplish knowledge-intensive tasks have garnered significant interest within database and AI communities. silos of multimodal data sources make it difficult to identify appropriate data sources for accomplishing the task at hand. We propose CMDBench, a benchmark modeling the complexity of enterprise data platforms.
  arXiv  Detail & Related papers  (2024-06-02T01:10:41Z)
• DACO: Towards Application-Driven and Comprehensive Data Analysis via Code Generation [83.30006900263744]
  Data analysis is a crucial analytical process to generate in-depth studies and conclusive insights. We propose to automatically generate high-quality answer annotations leveraging the code-generation capabilities of LLMs. Our DACO-RL algorithm is evaluated by human annotators to produce more helpful answers than SFT model in 57.72% cases.
  arXiv  Detail & Related papers  (2024-03-04T22:47:58Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.