Музыка • Comida • Teangacha • İş
I design real‑time rendering pipelines, programming models, and prototyping tools that stay fast under mobile constraints — latency, power, and memory.
Carnegie Mellon University, MS
Mumbai University, BS
Shri Bhagubhai Mafatlal Polytechnic
Apple, iOS/MACOS SWE
Walmart, iOS SWE
Paypal, iOS SWE
Two projects that best represent how I think: architecture‑aware performance, expressive tools, and real‑world interactive systems.
Restructuring multi‑pass pipelines to match tile execution and reduce device‑memory traffic.
A tile‑locality‑aware programming model that exposes residency semantics for compiler optimization.
Additional work across graphics tooling, interactive prototypes, and production systems. (Brief summaries; details available on request.)
A Swift-based machine learning framework for real-time computer vision on iOS.
A lightweight ray tracing pipeline tuned for iPad-class constraints.
A sketching and prototyping interface built on top of a custom real‑time graphics engine.
An interactive environment for learning and experimenting with GPU shader programming.
A simulator to visualize print paths and teach 3D printing concepts interactively.
An iOS travel app featuring an immersive 360° video player built with SceneKit.
A collection of iOS applications shipped to the App Store and used by thousands of users.
I’m interested in how carefully exposing architectural structure (like tile locality) can unlock safe, automatic optimization — and make interactive systems feel more predictable and “alive.”
Tile‑based GPUs execute fragments in on‑chip tile memory, yet many pipelines force intermediate writes to device memory. I explore restructuring passes so data stays tile‑local, reducing bandwidth, latency, and energy.
TileLang makes locality explicit through residency semantics. A compiler can then safely fuse passes and remove transient resources, producing efficient Metal pipelines while keeping code expressive for rapid iteration.
Got a chance to work at Carnegie Mellon University Labs
Developed and published a research iOS application in collaboration with Uber to study real‑world commuting patterns. Built the app end‑to‑end in Swift, integrated Uber APIs for route analysis, and designed the UI from scratch for clarity and usability.
Collaborated with the CMU Graphics Lab to build a real‑time simulation and visualization tool using C++ and OpenGL, focusing on efficient rendering pipelines and interactive performance for research workflows.
A small set of moments where I shared work publicly.
Interactive graphics + prototyping tool exploring performance‑aware visual systems for learning.
Teaching & Mentorship (Programming + Math)
Mentored students through project‑based learning in programming and mathematics, helping them design and build independent passion projects while developing strong computational foundations.
Mentored a student in designing and building a complete mobile application, providing guidance on architecture, implementation, and user experience. The project received Second Prize at the 2025 Congressional App Challenge, highlighting the effectiveness of project-based technical mentorship.
Consistent language learning through daily practice
Actively learning Spanish, Russian, and Irish through daily Duolingo practice. Maintaining a long‑term daily streak.
I’m a software engineer focused on real‑time graphics, performance, and interactive tooling (Swift, Metal, C++/OpenGL). I enjoy turning architectural constraints into design leverage — especially on mobile/embedded platforms.
