Ryan Vandersmith

Ryan Vandersmith is an R&D software engineer with over 10 years of experience and a deep self-taught background in meteorology. He builds weather forecasting and analysis tools used by photographers, storm chasers, and researchers, including Mesoview (skychaser.mesoview.app), a cloud inversion forecast tool, a mountain wave visualization tool, and a stepwise CAPE/CIN contribution to SounderPy. Professionally, he worked on Caffeine.ai, an AI website builder, and led Motoko programming-language tooling, including a VS Code extension with more than 85,000 installs. He also has experience with cross-chain EVM integrations, developer-experience products, UX research, Rust research software at the University of Colorado Boulder, consulting, and public-sector software.

Pandas GPT brings natural-language requests into a pandas workflow. Use `df.ask(...)` to transform data, answer questions with tables, or create visualizations, with options to preview generated code or allow changes to the original DataFrame.

The library supports configurable model backends, including OpenAI, Azure OpenAI, and LiteLLM providers such as Gemini, Hugging Face, and OpenRouter. Custom completers are also supported.

[GitHub repository](https://github.com/rvanasa/pandas-gpt) · [Medium article](https://levelup.gitconnected.com/pandas-gpt-power-up-your-data-science-workflow-with-ai-dataframe-queries-868515d798f0) · [PyPI package](https://pypi.org/project/pandas-gpt/) · [Google Colab demo](https://colab.research.google.com/github/rvanasa/pandas-gpt/blob/main/notebooks/pandas_gpt_demo.ipynb)

A Python library for querying, transforming, and visualizing pandas DataFrames with ChatGPT or other large language models.

PandasGPT