PandasAI
Developer(s)Gabriele Venturi, Massimiliano Pronesti, Arslan Saleem
Initial releaseApril 2023
Stable release
v2.0 / March 2024
Repositorygithub.com/Sinaptik-ai/pandas-ai
Written inPython and JavaScript
Typelibrary for conversational data analysis
LicenseMIT License
Websitepandas-ai.com

PandasAI is an open-source Python library that enhances the functionality of the popular pandas library by integrating it with Large Language Models (LLMs). This approach allows users to interact with data frames through natural language.[1], making data analysis more accessible and efficient[2]. PandasAI is distributed under MIT license.

How it works

edit

PandasAI provides a natural language semantic layer that uses large language models (LLMs) to convert the user query into pandas and SQL code[3]. The code is then executed and the output of the analysis is returned to the user, in the format of a pandas dataframe, a number, a string or a chart.

PandasAI is not limited to the pandas library, but it can use the whole ecosystem, including libraries like NumPy for computantion-heavy calculation, scikit-learn for basic ML tasks[4], matplotlib, seaborn and plotly for data visualization[5][6].

History

edit

PandasAI was launched on April 2023 as an open source project by Gabriele Venturi. It immediately earned traction, reaching the milestone of 5.000 stars on GitHub in approximately 2 weeks[7].

The company, based in Munich, after incorporating in May 2023 announced a $1.1M pre-seed round from Runa Capital, Episode 1 and Exor Ventures in September 2023[8][9][10].

In Q1 2024, the company has taken part of the Y Combinator W24 batch[11].

References

edit
  1. ^ "Pandas AI: Supercharging data analysis with generative AI". DEV Community. 2023-06-18. Retrieved 2024-03-10.
  2. ^ "Pandas AI: The Generative AI Python Library". GeeksforGeeks. 2023-06-08. Retrieved 2024-03-10.
  3. ^ "Pandas AI: The Generative AI Python Library". KDnuggets. Retrieved 2024-03-10.
  4. ^ "Official documentation". docs.pandas-ai.com. Retrieved 2024-03-11.
  5. ^ Muskan, Vivek (2023-07-01). "How to leverage AI for Data Visualization using pandas?". Medium. Retrieved 2024-03-10.
  6. ^ Walker, Michael (May 2024). Python Data Cleaning Cookbook: Prepare your data for analysis with pandas, NumPy, Matplotlib, scikit-learn, and OpenAI (2nd ed.). Packt. pp. 122–132. ISBN 9781803239873.{{cite book}}: CS1 maint: date and year (link)
  7. ^ "GitHub Star History". star-history.com. Retrieved 2024-03-10.
  8. ^ "PandasAI raises $1.1M in Pre-Seed round for AI-powered data analysis". Tech.eu. 2023-09-29. Retrieved 2024-03-10.
  9. ^ Escárzaga, Antonio L. (2023-09-29). "Munich-based PandasAI raises over €1 million for its AI-powered data analysis assistant". EU-Startups. Retrieved 2024-03-10.
  10. ^ "Why 'AI copilot' startups are so hot with VCs right now". Fortune. Retrieved 2024-03-11.
  11. ^ "PandasAI: PandasAI is an open-source conversational data analysis platform with…". Y Combinator. Retrieved 2024-07-10.