Saber Sheybani
Postdoctoral Researcher at Stanford University
Palo Alto, California, United States
Summary
👤
Senior🎓
Top SchoolSaber Sheybani is a Postdoctoral Researcher at Stanford with a PhD in AI and nine years of experience translating cutting‑edge machine learning research into production-ready systems for healthcare and vision. He specializes in deep representation learning, multimodal modeling, large-scale data pipelines, and LLM-powered clinical NLP, and has published at top venues including a NeurIPS Spotlight. At Stanford he built end-to-end pipelines on massive clinical datasets to produce early-risk models and a 300-cluster patient representation that revealed population-level health trajectories and treatment strategies tied to reduced opioid risk. His prior work on video and egocentric vision produced practical gains in self-supervised learning and scalable PyTorch training pipelines, and he led the ModelVsBaby benchmark to probe VLM failure modes. An active open-source contributor, he improved denoising methods in the widely used DIPY medical-imaging library, reflecting attention to numerical detail and reproducible tooling. He combines causal inference and agentic reasoning prototypes with a pragmatic drive to ship reliable, deployable AI systems in clinical and real-world settings.
9 years of coding experience
8 years of employment as a software developer
PhD Artificial Intelligence and Cognitive Science, PhD Artificial Intelligence and Cognitive Science at Indiana University BloomingtonBachelor of Science (B.Sc.) Electrical Engineering - Control Systems, Bachelor of Science (B.Sc.) Electrical Engineering - Control Systems at University of TehranHigh School Diploma Mathematics and Physics, High School Diploma Mathematics and Physics at NODET (Iran's National Organization for Development of Exceptional Talents) - Qazvin BranchPostdoctoral Training Biomedical AI - Entrepreneurship, Postdoctoral Training Biomedical AI - Entrepreneurship at Stanford University
English, Persian, German, Spanish
Github Skills (8)
machine-learning10
diffusion-mri10
python10
denoising10
numpy9
scipy9
medical-imaging8
testing8
Programming languages (3)
Github contributions (5)
dipy/dipy
Jun 2017 - Aug 2017
DIPY is the paragon 3D/4D+ medical imaging library in Python. Contains generic methods for spatial normalization, signal processing, machine learning, statistical analysis and visualization of medical images. Additionally, it contains specialized methods for computational anatomy including diffusion, perfusion and structural imaging.
Role in this project:
Data Scientist
Data ScientistContributions:20 commits, 2 PRs, 12 comments in 1 month
Contributions summary:Saber primarily contributed to the denoising module of the DIPY library. Their work involved implementing and refining the localPCA denoising method, including the addition of an eigenvalue decomposition (eigh) version. They made several changes to improve the function's functionality, such as making the mask optional, and corrected associated tests. The user's commits also included fixes for minor coding errors, which demonstrates a strong attention to detail.
signalpythonmicrostructurespatialtractography
ssheybani/sp17-i524
Mar 2017 - May 2017
Contributions:18 PRs, 32 pushes in 1 month
spring-bootjavaspringsubmissions
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.