This summer, I will join Lenovo's Qira team to build cross-device AI experiences across Windows and mobile platforms, with a focus on multimodal interaction, context handoff, and reliable user-facing agent behavior.
hello, i'm
Rohit Mahesh
I am a student at Columbia University studying computer science and mathematics in the school of engineering (SEAS).
I'm especially interested in machine learning and its applications, though more generally I just like building interesting things in software and tackling challenging problems.
The pattern on this page is Rule 110, a tiny Turing-complete cellular automaton, generated with SCSS and animated with pure CSS!
email github linkedin twitter portfolio
next
research
I developed a longitudinal medical-imaging pipeline for tuberculosis outcome prediction, extracting Vision Transformer representations from serial chest X-rays and modeling temporal disease progression with sequence-level aggregation.
I build machine-learning and scientific-computing tools for quantifying uterine contractility from microscopy data. My projects include Utero-ViT, a Vision Transformer model for classifying contractility dynamics, and WaveAtlas, a web platform for extracting, visualizing, and analyzing wave-propagation features from high-dimensional biological time series.
engineering
I developed software for multiline tactile Braille displays, including an English-to-Braille translation engine, an accessible menu framework for tactile navigation, and low-latency audio cursor cues iterated with blind co-designers.
patents + publications
- Pending patent: Low-cost refreshable Braille display (US 63/706,318).
- Imaging the dynamics of murine uterine contractions in early pregnancy . Biology of Reproduction, 2024.
- The effects of dysregulated ion channels and vasoconstriction in glioblastoma multiforme . Journal of Emerging Investigators, 2025.
elsewhere
I am also interested in applied security, math competitions, and robotics.
In my free time I can be found playing horror video games or deep-diving into my OS stack (fedora)