Agrin Hilmkil is a research engineer with 12 years of experience building and deploying ML systems, currently teaching Gemini to become a better agent at Google DeepMind. He has driven inference, product and infrastructure work for Latent Labs’ Latent-X and written high-performance CUDA kernels for quantized transformers at Microsoft, blending research depth with production-focused engineering. His background spans speech and generative models—leading a team that shipped a best-in-class Swedish TTS stack and an entirely synthesized audiobook—plus applied physiological RNN research from academia. Based in London, he combines strong mathematics and engineering training from Chalmers with a knack for turning experimental models into reliable inference pipelines. Notably, his work ethos—playfully summarized as “overfitting on test sets since 2012”—reflects a hands-on, iterative approach to squeezing performance from models in real-world settings.
12 years of coding experience
12 years of employment as a software developer
Combined program with BSc and MSc, BSc in Engineering Physics and MSc in Engineering Mathematics and Computational science, Combined program with BSc and MSc, BSc in Engineering Physics and MSc in Engineering Mathematics and Computational science at Chalmers University of Technology
Contributions:6 commits, 5 PRs, 5 pushes in 1 year 5 months
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
Agrin Hilmkil - Research Engineer at Google DeepMind