Sergio Perez

I’m passionate about building AI- and math-driven software to tackle complex real-world problems. My experience spans cutting-edge AI research, simulation, and software development — across both academia and industry — alongside initiatives in open education, science policy, and communication aimed at driving positive impact. Currently, I’m a Solutions Architect for AI at NVIDIA, serving as a technical bridge between NVIDIA and developers, researchers, and data scientists. I design proof-of-concept solutions, advise on AI acceleration, and help organizations adopt GPU-powered computing. Previously, I was an AI Engineer at Graphcore (2021–2023), optimizing state-of-the-art models (GPT, BERT, ResNet50, GNNs) on Graphcore’s IPU architecture using PyTorch and TensorFlow. I earned my PhD at Imperial College London (2021), where I developed numerical methods for data-driven fluid dynamics — research recognized through international collaborations, conferences, and publications. Earlier, I worked as a Machine Learning Scientist Intern at Amazon (2020), leading automation and forecasting projects for the UK Supply Chain team during the COVID-19 crisis.