Shahbuland Matiana is Head of Mathematical Research with seven years of hands-on experience building and optimizing generative models for video, compression, and immersive worlds. He has led research and engineering efforts across startups and labs—including Stability AI, SynthLabs, and his current role at Overworld—focusing on diffusion world models, deep compression autoencoders, and fast video-data pipelines. Shahbuland combines applied math and systems thinking to squeeze performance from large models and real-time pipelines, and has shipped tooling for RLHF on both language and diffusion models. Based in Canada and trained in Data Science at the University of Waterloo, he works toward AI-generated games and full-dive VR with a practical taste for scalable synthetic data and multimodal simulators. Notably, his background spans both foundational model research and production-focused libraries (tRLX/DRLX), reflecting a rare mix of research depth and deployment fluency. His GitHub playfully brands AGI as “Artificial Goose Intelligence,” hinting at a pragmatic, creative approach to ambitious AI goals.
7 years of coding experience
4 years of employment as a software developer
Bachelor's degree, Data Science, Bachelor's degree, Data Science at University of Waterloo
magiCARP is an API used for crossencoder training.
Contributions:3 reviews, 97 commits, 17 PRs in 3 months
nlpapideep-learningmachine-learningtraining
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
Shahbuland Matiana - Head Of Mathematical Research