top of page
ai lightspeed uncroppedfinal.png

Thanks! We'll contact you with important updates!

Logo with VMware.png
Logos with new hp.png

Find out why we're trusted by many organisations.

Goodbye to waiting times

Moonshot's algorithms are proven to be the fastest in the world as confirmed by many peer-reviewed research papers.


Our algos run 2000x faster, use 50% less resources, and they work on all devices. Our first product launch Unsloth is just the beginning


AI models are getting larger and more power hungry. Slow 3 hour predictions hamper our ability to make life changing decisions.

With 1.5K Github stars and used by 20+ huge organizations, Hyperlearn powers our simulation models and makes AI faster for the world.

Making AI faster and greener

Image by Markus Spiske
Image by Igor Son

Making Classical Machine Learning Pipelines Differentiable: A Neural Translation Approach

"Hyperlearn provides a Scikit-learn interface with optimized algorithms for linear/ridge
regression, PCA, SVD, etc."

Fractional ridge regression: a fast, interpretable reparameterization of ridge regression

"This functionality relies on an implementation provided in the Hyperlearn library and copied into the fracridge source-code"

ZikaHack 2016: A digital disease detection competition

"The solution produced signals of possible ZIKV outbreak as early as September 2014 in the state of Espírito Santo."

Survey of Text-based Epidemic Intelligence: A Computational Linguistics Perspective

Citation reference 2: Dillon C. Adam, Jitendra Jonnagaddala, Daniel Han-Chen, Sean Batongbacal, ... et al. 2017.

We've launched our first product: Unsloth!

Join our subscription list so you'll be notified of the most important updates!

Thanks! We'll contact you when we launch!

bottom of page