Summary
Sagar Shrestha is an applied scientist with 9 years of engineering and research experience building scalable ML systems spanning generative AI, graph learning, and multimodal models. He has 12+ top-tier publications (NeurIPS, ICLR, ICML, TSP, ICASSP) and a track record of translating novel research into production at Amazon, Samsung Research, and in a robotics/AI startup he co-founded. His work includes practical wins—such as a semi-supervised heterogeneous graph transformer that improved ATO detection at scale—and foundational contributions like identifiability-guaranteed unsupervised domain translation and communication-efficient federated multiview learning. Comfortable across the stack, he has led end-to-end product deployments from perception and planning for service robots to production chatbots and backend systems. Based in Corvallis, OR, he pairs rigorous PhD-level research with hands-on engineering pragmatism, often optimizing for inference speed and deployment constraints in real-world settings.
9 years of coding experience
4 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, 4.0/4.0, Doctor of Philosophy - PhD, Computer Science, 4.0/4.0 at Oregon State University
Engineer’s Degree, Electronics and Communication Engineering, Engineer’s Degree, Electronics and Communication Engineering at Institute of Engineering, Pulchowk Campus