Summary
Akshay Mehra is an ML researcher with a Ph.D. in Computer Science from Tulane University and a decade of experience spanning research and industry, now working at Dolby Laboratories. His work rigorously studies how distribution shifts impact accuracy, adversarial robustness, and transferability, and he builds algorithms informed by those theoretical insights to improve real-world model performance. He has interned at Apple and Lawrence Livermore National Laboratory and brings a rare blend of academic depth and applied engineering from prior roles at Microsoft and research positions. Akshay’s research uniquely connects in-domain, cross-domain, and cross-task evaluation, revealing practical failure modes of models under distribution drift. Based in Atlanta, he combines top academic credentials (OSU MS, high honors BE) with a demonstrated track record of translating theoretical findings into deployable ML solutions.
10 years of coding experience
7 years of employment as a software developer
High school, High School, High school, High School at Delhi Public School, Mathura Road
Bachelor of Engineering - BE, Computer Science, CGPA: 9.71/10, Bachelor of Engineering - BE, Computer Science, CGPA: 9.71/10 at Thapar Institute of Engineering and Technology
Doctor of Philosophy - PhD, Adversarial Machine Learning, Robustness to distribution shifts., Doctor of Philosophy - PhD, Adversarial Machine Learning, Robustness to distribution shifts. at Tulane University
Master of Science - MS, Computer Science, GPA: 4/4, Master of Science - MS, Computer Science, GPA: 4/4 at The Ohio State University