This project provides an LLM-powered system for automatically recommending and generating interactive visualizations from data. It targets users who need to quickly create charts without deep expertise in data visualization, offering an efficient way to transform data into interactive HTML outputs. The system's core benefit lies in lowering the barrier to chart selection and improving visualization workflow efficiency through intelligent recommendations and a streamlined generation process.

How It Works

VizPilot AI accepts data uploads in CSV or Excel formats, automatically profiling fields and analyzing data characteristics. It then leverages LLMs (supporting DeepSeek, MiniMax, and OpenAI-compatible APIs) to recommend 3-5 suitable chart types, complete with rationale and suggested field mappings. Users select a recommended chart, and the system generates an interactive Plotly-based HTML visualization. A key architectural choice is the registry-driven chart system, where chart metadata is managed centrally in charts/registry.py, allowing new chart types to be added easily without altering core logic, ensuring front-end and back-end synchronization.

Quick Start & Requirements

• Primary install / run command:
  • Install dependencies: pip install -r requirements.txt
  • Run the application: python app_pro.py (or start.bat on Windows)
• Non-default prerequisites and dependencies: Python 3.7+, pip. API keys for chosen LLM providers (DeepSeek, MiniMax, or OpenAI-compatible) are required for the recommendation feature.
• Access: Open http://localhost:5017 in a browser.
• Links: Screenshots are available in the README.

Highlighted Details

• Supports CSV and Excel (.xlsx) file uploads with automatic field type identification (numeric, categorical, date).
• Integrates with multiple LLM providers (DeepSeek, MiniMax, OpenAI-compatible) for chart recommendations.
• Features a registry-driven architecture for managing over 45 supported chart types, enabling easy extensibility.
• Generates interactive Plotly HTML charts with download support.

Maintenance & Community

The project includes a roadmap outlining future development, such as automatic chart parameter tuning, multi-chart dashboarding, and SQL data source integration. No specific community channels (like Discord or Slack) or notable contributors/sponsorships are mentioned in the README.

Licensing & Compatibility

The project recommends using the GPL License. As a copyleft license, GPL may impose restrictions on integrating this software into closed-source or proprietary commercial products without making the derived work's source code available under the GPL.

Limitations & Caveats

The system requires LLM API keys for its core recommendation functionality. While over 45 chart types are listed, the definitive source is charts/registry.py, suggesting the list may evolve. Future development items on the roadmap indicate that features like automatic parameter tuning and dashboard creation are not yet implemented.