Pramith Devulapalli

About Me

I’m on the 2025–2026 job market and actively seeking research scientist and other research-related positions.

I am a Ph.D. candidate in Computer Science at Purdue University and a graduate researcher at Sandia National Labs.

The core of my Ph.D. focuses on the theoretical foundations of machine learning where I have the privilege of being advised by Prof. Steve Hanneke. I am also a graduate researcher funded by Sandia National Labs where I am fortunate to work with Prof. Ananth Grama and Dr. William Hart. Additionally, I am grateful to be supported by the Purdue Doctoral Fellowship earlier in my Ph.D. Previously, I obtained my B.S. in Engineering Physics from Case Western Reserve University. During my undergrad, I was a bioinformatics researcher where I was fortunate to be advised by Prof. Satya Sahoo. Prior to undergrad, I worked in condensed-matter physics under Prof. Andres La Rosa, to whom I’m grateful for my first research opportunity.

Research Interests

Broadly speaking, my research lies within machine learning theory, where we seek to establish theoretical guarantees on the performance and efficiency of learning algorithms across a variety of settings. A central part of my work is online learning, which studies how an algorithm performs as it receives data one-by-one in a streaming fashion. In short, I ask when and how a learning algorithm can adapt reliably as data arrives over time. I’m particularly interested in extending online learning to continuous data streams, capturing phenomena that evolve at every instant rather than at discrete steps, and in understanding what constitutes learnability in this setting.

Papers

Publications

• Uniform Convergence Beyond Glivenko-Cantelli
  Tanmay Devale, Pramith Devulapalli, and Steve Hanneke. To appear in ALT, 2026.
• Agnostic Continuous-Time Online Learning
  Pramith Devulapalli**, Changlong Wu**, Ananth Grama, and Wojciech Szpankowski. NeurIPS 2025.
• Learning from Snapshots of Discrete and Continuous Data Streams
  Pramith Devulapalli and Steve Hanneke. NeurIPS 2024.
• The Dimension of Self-Directed Learning
  Pramith Devulapalli and Steve Hanneke. ALT 2024.
• Embedding Conjugate Gradient in Learning Random Walks for Landscape Connectivity Modeling in Conservation
  Pramith Devulapalli*, Bistra Dilkina, and Yexiang Xue. IJCAI 2020.
• Computing functional brain connectivity in neurological disorders: efficient processing and retrieval of electrophysiological signal data
  Arthur Gershon, Pramith Devulapalli*, Bilal Zonjy, Kaushik Ghosh, Curtis Tatsuoka, and Satya S. Sahoo. AMIA 2019.
• Instrumentation-level improvements in shear-force near-field acoustic microscopy
  Jianghua Bai, Pramith Devulapalli*, Theodore Brockman, and Andres H. La Rosa. Journal of Physics: Conference Series, 2018.