DataHeroes is building a Python library that solves some of today’s biggest machine learning challenges: reducing the time, effort and resources required to develop and maintain high-quality machine learning models.
DataHeroes uses a unique methodology called Coresets, originating in computational geometry, to reduce the size of the dataset to a small subset that maintains the statistical properties and corner cases of the full dataset, such that solving a problem on the Coreset will yield the same result as solving it on the full dataset (to learn more about Coresets, visit our Introduction to Coresets page).
Reducing the dataset size creates huge advantages:
We’ve assembled a world class team of engineers, data scientists and coreset researchers and teamed up with world class investors to help us bring our dream to fruition.
Achieving even some of these lofty goals will have monumental effects on the advancements of AI in a responsible and environmentally conscious way in the coming years and decades and we are extremely excited to be able to play even a small role in that and making our contribution to the data science community and to the world!