ruc-datalab/DeepAnalyze

DeepAnalyze is the first agentic LLM for autonomous data science. 🎈你的AI数据分析师，自动分析大量数据，一键生成专业分析报告！

What it solves

DeepAnalyze is designed to automate the entire data science pipeline, removing the need for human intervention in tasks like data preparation, analysis, modeling, visualization, and report generation. It specifically addresses the challenge of conducting open-ended data research across diverse data sources, including structured (Databases, CSV, Excel), semi-structured (JSON, XML, YAML), and unstructured (TXT, Markdown) formats.

How it works

DeepAnalyze is an agentic Large Language Model (LLM) that can autonomously explore data sources and execute data-centric tasks. It is provided as an 8B parameter model (DeepAnalyze-8B) that can be deployed via vLLM. The system supports various interfaces for interaction, including a WebUI, a JupyterUI (which converts LLM outputs into executable code cells), and a CLI. It can be integrated into workflows via an OpenAI-style API.

Who it’s for

This project is for data scientists, researchers, and analysts who want to automate their data analysis workflows and generate analyst-grade research reports from raw data files.

Highlights

  • End-to-End Automation: Handles everything from data preparation to final report generation.
  • Multi-Format Support: Works with structured, semi-structured, and unstructured data.
  • Open Source: The model, training data (DataScience-Instruct-500K), and code are all fully open-sourced.
  • Flexible Deployment: Supports multiple interfaces including WebUI, JupyterUI, and CLI, and can be deployed using vLLM with various quantization options for different GPU memory constraints.

Related

  • Project
    FireBird-Technologies/Auto-Analyst

    An open-source AI data scientist that automates data cleaning, statistical analysis, and machine learning via a modular multi-agent system.

  • Project
    confident-ai/deepeval

    An open-source LLM evaluation framework that provides unit-testing capabilities for AI agents, RAG pipelines, and chatbots using LLM-as-a-judge metrics.

  • Project
    416rehman/DeepZero

    An automated vulnerability research pipeline engine that uses YAML-defined workflows to orchestrate data ingestion, filtering, and LLM-based security assessments.

  • Project
    eosphoros-ai/DB-GPT

    An open-source agentic AI data assistant that connects to databases and files to autonomously write SQL and code for data analysis and reporting.

  • Project
    microsoft/RD-Agent

    An autonomous multi-agent framework that automates the research and development loop of proposing and implementing models and data features for ML engineering and quantitative finance.