About me

Hi! I’m Pranati Modumudi, a first-year Master’s student studying Computer Science at Columbia University, specializing in the Machine Learning and Neuroscience track. I completed my undergraduate studies at UC Berkeley, where I studied Data Science (with an emphasis in Sociology) and Economics and was a recipient of the Regents’ and Chancellor’s Scholarship at Cal.

I’m driven by a deep curiosity about intelligence, both natural and artificial, and how insights from each domain can elevate and inform the other. My passion lies in building ethical, responsible, and inclusive AI systems that are transparent, safe, and aligned with human values. I care deeply about designing technology that not only works — but works for everyone.

I’m currently involved in three projects spanning my core interests:

• NeuroAI & Continual Learning: MS Thesis advised by Professor Rich Zemel exploring biologically-inspired learning and memory mechanisms to build adaptive and resilient ML systems.

• Brain-Computer Interfaces & Decision Neuroscience: Studying human decision-making under risk and ambiguity using EEG and eye-tracking at Columbia’s Columbia’s Laboratory for Intelligent Imaging and Neural Computing, developing pipelines to model behavior from noisy neural signals.

• AI Safety & Alignment: Investigating robustness, failure modes, and pluralistic approaches to AI alignment in collaboration with Glen Weyl, focusing on how diverse human values can be better represented in AI.

Previously, I spent nearly two years as a Data Scientist at Disney Streaming, building production ML pipelines for causal inference and time-series modeling that affected millions of subscribers. At Berkeley, I conducted research at the Redwood Center for Computational Neuroscience.

Outside school, I love dancing, boxing, and exploring new cuisines.

Currently Reading

• 1q84, Haruki Murakami
• The Left Hand of Darkness, Ursula K. Le Guin
• The Precipice: Existential Risk and the Future of Humanity, Toby Ord