Welcome!

I am Deepak, a Ph.D. candidate at Purdue Univeristy’s Neuro(nano) Research Laboratory advised by Prof. Kaushik Roy. My research focuses on trustworthy AI.

About me

Deepak is currently pursuing his Ph.D. under the guidance of Prof. Kaushik Roy at Purdue University, where his research has gained significant recognition. His work has been spotlighted at leading conferences, earning the 2024 NeurIPS Spotlight Paper Award (Top 2%), the 2024 ICML Spotlight Paper Award (Top 3.5%) and Estus H. and Vashti L. Magoon Research Excellence Award 2025. At Purdue, he has been awarded the College of Engineering Scholarship and the ECE Summer Research Grant. His research focuses on deep learning algorithms, with a particular emphasis on deep learning memorization and trustworthy machine learning.

He previously worked as an ML Research intern at Microsoft, where he focused on predictive time-series machine learning models. Prior to that, at National Instruments R&D, he developed innovative signal acquisition and processing frameworks, with his work being recognized as one of the Top 3 Best Papers at the National Instruments Tech Conference in 2017.

Deepak earned his M.S. in Electrical and Computer Engineering from Purdue University in 2019, where he received the prestigious Magoon Teaching Excellence Award for his outstanding contributions as a teaching assistant. He completed his B.E. in Electronics Engineering from M. S. Ramaiah Institute of Technology, India, in 2016, graduating as a bronze medalist for academic excellence.

Deepak Ravikumar

  • (2025 exp.) Ph.D. Computer Engineering, Purdue University
  • (2019) M.S. Computer Engineering, Purdue University
  • (2016) B.Tech ECE, M.S. Ramaiah Institute of Technology

News

20 June, 2025

Defended my PhD research, Towards Trustworthy AI: Understanding Memorization, Privacy, and Security in Deep Learning. See pictures on LinkedIn. You can see my defense on YouTube.

31 May, 2025

Authored invited paper on Building Resilient AI: Strengthening Data, Security, and Robustness in Neural Networks at IOLTS 25.

1 May, 2025

Towards Memorization Estimation: Fast, Formal and Free, accepted at ICML '25

... see all News