Jay Roberts

Member Of Engineering at poolside

Stony Stratford, England, United Kingdom
email-iconphone-icongithub-logolinkedin-logotwitter-logostackoverflow-logofacebook-logo
Join Prog.AI to see contacts
email-iconphone-icongithub-logolinkedin-logotwitter-logostackoverflow-logofacebook-logo
Join Prog.AI to see contacts

Summary

👤
Senior
🎓
Top School
Jay Roberts is an AI-focused engineer with a Ph.D. in Mathematics and eight years of experience building and scaling research-driven ML systems across academia, government labs, and industry. He blends deep mathematical theory with practical engineering—developing privacy guarantees for LLMs, new loss functions that cut training time by over 40%, and large-scale distributed training and deployment at Amazon and MIT Lincoln Laboratory. Currently a Member of Engineering at poolside, he focuses on core evaluations of AI models while carrying a track record of translating theoretical results to product and customer-facing explanations. His background includes novel contributions to explainable AI, uncertainty-aware training, and seismic inversion using stochastic dynamical systems, reflecting a strong multidisciplinary bent. Based in Seattle, Jay brings rare comfort moving from formal proofs to multi-GPU production fine-tuning and stakeholder-facing presentations. An early-career pattern is his ability to shorten iteration cycles without sacrificing model guarantees—turning mathematical insights into measurable engineering wins.
code9 years of coding experience
job4 years of employment as a software developer
bookPh.D Mathematics, Ph.D Mathematics at UC Santa Barbara
languagesSpanish
github-logo-circle

Github Skills (16)

pytorch-lightning3
yaml-configuration3
neural-network3
reproducible-science3
yaml2
dataclasses2
scalable2
machine-learning2
deep-learning2
pytorch2
optuna1
crossword1
tensorflow1
robustness1
optimization1

Programming languages (2)

Jupyter NotebookPython

Github contributions (5)

github-logo-circle
KyleMylonakis/GeoSeg

Aug 2018 - May 2019

Implementations of various neural network architectures to infer the interface depth of simulations of the elastic wave equation with the frozen gaussian approximation.
Contributions:116 commits, 4 PRs, 44 pushes in 8 months
neural-network
Contributions:94 pushes, 8 branches in 6 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