I grew up in motion. Six cities across India, each one a different language at the dinner table, a different school uniform, a different version of home. By the time I was sixteen, I had attended eight schools. Some people find roots in places. I found mine in the act of leaving them.
Then came UWC Mahindra, a boarding school in the hills of Pune where sixty-plus nationalities collided under one roof. It cracked open the world for me. I sat across from people whose childhoods were shaped by conflicts I had only read about, whose governments had failed them in ways I could barely fathom. And yet, the kid from Palestine and the kid from a small town in Bengal found the same humor in bad cafeteria food. That was the first time I understood that inequity is not a statistic. It is the gap between what you see at home and what you see everywhere else.
I came to Vanderbilt not to escape that gap, but to learn how to close it.
Now, as a BS/MS student in Computer Science, Economics, and Applied Mathematics, I work at the intersection of mathematical theory and real-world systems. At IBM watsonx AI Labs, I build self-directed AI agents for live marketplaces. At Vanderbilt's Center for Transportation, I apply reinforcement learning and agent-based simulation to model Nashville's traffic corridors, optimizing congestion strategies across 10,000+ traffic patterns using R-tree indexing and dynamic programming. In the Advanced Learning Lab, I engineer multimodal ML pipelines to predict student metacognition with 88% accuracy.
My current research explores training neural networks on fractal geometries, investigating how self-similar mathematical structures can inform more efficient network architectures. It sits at the intersection of pure mathematics and deep learning, the kind of problem that demands both rigorous theory and creative experimentation.
I think in math and build in code. Whether it is domain-adaptive pre-training for NLP models, LiDAR-guided autonomous navigation for search-and-rescue drones, or GPU-accelerated deep learning workflows for visual cognition, I gravitate toward problems where computational rigor meets tangible impact. I also play the violin and write, because some of the best ideas arrive outside the terminal.
I see technology as infrastructure. The systems that matter most are the ones built for communities that have historically been excluded from the design process.