Sam Devlin
AI Research Scientist at Meta
Stony Stratford, England, United Kingdom
Summary
👤
Senior🎓
Top SchoolSam Devlin is an AI research scientist with a decade of experience advancing multi-agent reinforcement learning, imitation learning, and multimodal generative models across academia and industry. After a PhD focused on multi-agent RL, he led game-AI research and secured substantial funding as a Research Fellow before joining Microsoft, where he progressed to Principal Researcher and tech-led cross-Atlantic v-teams delivering ML solutions for gaming and robotics. Now at Meta, he continues to build autonomous agents capable of sequential decision-making, combining deep research rigour with product-focused collaboration. Sam is as comfortable mentoring and managing teams—having supervised postdocs and PhD students and completed leadership training—as he is publishing and engineering novel algorithms. Based in Stony Stratford, UK, he blends academic depth with applied impact across large-scale industrial research programs. An understated thread through his career is a long-standing focus on game-derived problems as a proving ground for real-world multi-agent and sequential decision-making systems.
10 years of coding experience
13 years of employment as a software developer
Agile Project Management, Agile Project Management at McKenna ConsultantsMasters of Engineering Computer Systems and Software Engineering, Masters of Engineering Computer Systems and Software Engineering at University of YorkMachine Learning Summer SchoolIncluding Situational Leadership II & The Coaching Habit, Including Situational Leadership II & The Coaching Habit at Microsoft Internal Management Training
Github Skills (14)
bottleneck9
noise9
generalization8
ntt7
injection7
sum5
markov5
reinforcement-learning4
sam4
image-synthesis2
player2
convolutional-neural-networks2
deep-learning2
synthesis2
Programming languages (3)
Github contributions (5)
microsoft/IBAC-SNI
Oct 2019 - Oct 2019
Code to reproduce the NeurIPS 2019 paper "Generalization in Reinforcement Learning with Selective Noise Injection and Information Bottleneck" by Maximilian Igl, Kamil Ciosek, Yingzhen Li, Sebastian Tschiatschek, Cheng Zhang, Sam Devlin and Katja Hofmann.
Contributions:4 commits in 9 days
generalizationinjectionselectivereinforcement-learningneurips
microsoft/NTT
Jul 2021 - Nov 2022
Navigation Turing Test (NTT): Learning to Evaluate Human-Like Navigation [ICML 2021]
Contributions:3 commits, 3 pushes in 1 year 4 months
human-liketuring-testevaluateturingntt
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.