Summary
Sohrab Ferdowsi is an Associate Research Fellow and machine learning engineer with 9 years of experience designing low-latency, medtech-compliant image processing and computer vision pipelines for clinical devices and large-scale visual search systems. He translates ophthalmic imaging into actionable biomarkers at EarlySight and extracts clinically relevant insights across heterogeneous healthcare data, combining deep learning, GNNs and NLP with rigorous pen-and-paper ML foundations. His PhD work on learned compression and similarity-based retrieval informs practical, git-driven implementations in Python and C++, and he has led Innosuisse-backed projects spanning de novo molecular design, clinical trial risk prediction, and anomaly detection for product authentication. A proponent of bottom-up, simplicity-first ML deployments, he focuses on explainability and real-world risk management rather than chasing massive compute, and remains a vocal critic of prevailing AI narratives.
9 years of coding experience
4 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at University of Geneva
Master of Science (M.Sc.), Electrical Engineering- BioElectrics & Signal Processing, Master of Science (M.Sc.), Electrical Engineering- BioElectrics & Signal Processing at Sharif University of Technology
English, French, Persian