The name "Waraqa" is inspired by the Arabic word for "paper" or "leaf," symbolizing the simplicity of converting data to insights with AI and the lightweight nature of data analysis with Pyodide. The current version is a work in progress, with plans to add more features such as support for more file types, robust response handling, function calling for Ollama, and more. Contributions to the project are welcomed.
Key takeaways:
- Waraqa is an AI-powered data analysis tool that runs privately on your machine, offering features like Pyodide analysis environment, interactive visualizations, flexible model selection, and the ability to edit and re-execute code.
- The tool can be manually built on macOS by cloning the repository and running specific commands. However, its compatibility with Windows is yet to be tested.
- The name "Waraqa" is inspired by the Arabic word "وَرَقَة" (waraqa), which translates to "paper" or "leaf," symbolizing the simplicity of converting data to insights with AI and the lightweight nature of doing data analysis with Pyodide.
- While the tool is functional, there are several features and improvements in the pipeline, including support for more file types, robust response handling, support function calling for Ollama, support multi-cell responses and markdown outputs, visualization edits, and persistence of chats and chat histories on the local filesystem.