Summary
Hitesh Kumar is an AI Systems Engineer based in London with 9 years of experience designing, benchmarking and scaling datacentre‑scale AI and HPC clusters. He bridges low‑level performance work—writing optimized CUDA kernels and Fortran/PDE solvers—with higher‑level orchestration of thousands‑node deployments, benchmarking, and containerised IaaS for large‑scale AI training. His background includes performance engineering for accelerators (IPU/NVIDIA), building metrics and automation stacks, and practical experience porting and sizing compute, storage and networking with major vendors. Notably, he has applied numerical weather prediction expertise to real‑world forecasting and anomaly detection projects, illustrating a rare mix of scientific computing and production AI infrastructure. He writes about AI/HPC hardware and runs a focused infrastructure blog and community, aiming to shift attention from models to the machines that run them.
9 years of coding experience
6 years of employment as a software developer
MSc, ACSE (Computational science and engineering), MSc, ACSE (Computational science and engineering) at Imperial College London
English