SophoScale

Smarter models with less data

Mission

At SophoScale, we are committed to transforming the efficiency and cost-effectiveness of deep learning processes. Our focus is on developing advanced solutions that optimize GPU utilization during both the training and inference stages of model development. Here’s how we make a difference:

Reducing GPU Costs: By enhancing the processing speed and efficiency of GPUs, we dramatically decrease the expenses associated with running complex AI models.
Increasing Speed: Our proprietary technologies accelerate the training and inference phases, enabling faster model iterations and quicker deployment.
Data-Efficient ML: We leverage cutting-edge research in data-efficient machine learning to ensure that our models perform optimally with less data, reducing the resource footprint and improving outcomes.

Our solutions are designed for scalability and impact, helping organizations of all sizes implement more efficient AI without compromising on performance.

Founder

Veljko Kovac has extensive experience in AI modeling, holding an MSc in Artificial Intelligence from the University of Amsterdam where he received ELLIS Honours and cum laude distinction. His pioneering research on Graph Neural Networks was published at top-tier conferences like ICML.

He was a Visiting Researcher at the University of Cambridge, supervised by Dr. Veličković (Google DeepMind) and Prof. Dr. Pietro Liò (U. Cambridge). Before this, Veljko developed several AI models that are now deployed across all public hospitals in Hong Kong, significantly aiding daily medical practices.

TRUSTED BY

For further inquiries:

Page updated

Report abuse