Photo by Stefan Coders on Pexels

Plexe, a newly released open-source AI agent, is designed to simplify machine learning model creation. By leveraging natural language prompts, Plexe allows users to transform structured data, such as CSV and Parquet files, into functional ML models without requiring in-depth data science knowledge. This innovative tool functions as a team of specialized agents powered by large language models (LLMs), interpreting user descriptions to automatically generate the necessary models. The project, announced on Reddit (https://old.reddit.com/r/artificial/comments/1kkag85/we_built_an_opensource_ml_agent_that_turns/), with its source code available on GitHub (https://github.com/plexe-ai/plexe) and a walkthrough provided on YouTube (https://www.youtube.com/watch?v=bUwCSglhcXY), aims to democratize ML development. Future plans include expanding data source compatibility with database support and incorporating advanced feature engineering capabilities.