GPU Data Profiler 2025 
A Hackathon Project - Celebrating the Unappreciated Work of Machine Learning
Overview
This project represents the often overlooked but critical foundation of machine learning: data profiling and preprocessing. While flashy models get the spotlight, the real heroes are the data engineers and scientists who ensure clean, well-understood data flows through ML pipelines.
Vision
Our vision is to develop a dataframe library, inspired by NVIDIA’s RAPIDS, to perform computations on the GPU.
Numerical Analysis
Currently, we have implemented a GPU-optimized MAX
function.