Karl Marrett is a Member of Technical Staff specializing in optimizing ML backends for learning, search, and generation, with 11 years of experience building high-throughput, data-intensive systems across startups and academia. He tunes GPU kernels, systems, and infra to accelerate researcher iteration, applying HPC/CUDA to projects that contributed to Nature Neuroscience–cited advances in spatial AI. At UCLA he scaled brain-focused models dramatically (67x scale, 509x memory reduction) and designed a CPU+GPU‑optimized 3D vision pipeline (Gossamer) that earned recognition for performance and engineering. His work blends deep systems engineering with practical research support—freeing labs and developers to move faster in production-like settings. Based in Seattle, he brings a rare combination of academic rigor (PhD-level research) and production-first engineering that consistently turns challenging ML workloads into tractable, deployable systems.
11 years of coding experience
9 years of employment as a software developer
Bachelor of Science (B.Sc.), Computational Neuroscience, Applied Math, Bachelor of Science (B.Sc.), Computational Neuroscience, Applied Math at University of Washington
Contributions:20 pushes, 4 branches in 2 years 3 months
sparsedata-structurevolumeopenvdb
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.
Request Free Trial
Karl Marrett - Member Of Technical Staff at Reka AI