Welcome, I am Monishwaran.

I am a master's student in Electrical Engineering and Computer Sciences (EECS) at the University of California, Berkeley, where I conduct research in Artificial Intelligence. As an undergraduate, I earned dual degrees with Honors in Electrical Engineering and Computer Sciences (EECS) and Mathematics.

My research interests lie in the intersection of High Performance Distributed Computing, Large Scale Machine Learning, and High Dimensional Statistics. In the past, I have worked on research projects that explore these ideas independently, though my desire is to incorporate the aforementioned ideas together to create efficient inference and training of Large Neural Network Models.

At UC Berkeley, I am part of Berkeley AI Research (BAIR: PALLAS) focusing on optimizations of Large Language Models (LLMs) under the guidance and mentorship of Professor Kurt Keutzer. The projects that I am involved in span developing training infrastructure on GPUs and TPUs to train LLMs, improving the context length, inference speed, and memory footprint of LLMs, and building smart AI Agents through RAG and efficient function calling. Through the Lawrence Berkeley National Laboratory, I am involved in neutrino trajectory predictions (Nuclear Fusion research) using SOTA GNNs where I am advised by UC Berkeley Particle Physicist Yury Kolomensky and and Ph.D. student Tong Zhu. I am also involved in solving Partial Differential Equations (PDEs) using deep neural networks (DNNs), a project led by Professor Federico Pasqualotto.

In the past, I have worked on Neural Network Theory through the EECS Honors Program, under the supervision of Professor Federico Pasqualotto. This research laid the framework for my honors thesis—Scaling and Training Dynamics of Wide Neural Networks–which was later compiled and submitted as part of my honors degree in mathematics. Another notable project I engaged in involved Tensor Train Statistics, where I worked with Math Ph.D. student Michael Kielstra to build a High Performance Tensor Algorithms Library as part of the UC Berkeley Math Directed Reading Program.

Besides my academic work, I love to teach, and I share my projects and knowledge through numerous international keynotes and TEDx talks. In addition, out of pure interest, I spend time analyzing the works of Friedrich Nietzsche, Ludwig Wittgenstein, Bertrand Russell. When the weather is right, I surf in San Francisco. I also row at the California Lightweight Rowing Club at UC Berkeley.

My ultimate goal is to bridge the gap between disjoint academic fields and aspire to develop a unique expertise that not only solves complex problems but also creates a new area of research. I welcome collaboration and meeting people from different domains, and if you are interested in contacting me, feel free to email me at monishwaran@berkeley.edu

This website contains information from my research ventures, projects, and life stuff.