Hey, I’m Aditya 👋

I love building cool projects with AI, exploring new ideas, and sharing my travels 🌍✈️

About Me

I’m Aditya Krishna Ponnudurai, a Graduate Computer Science student at the University of Illinois Urbana-Champaign, passionate about building intelligent systems that bridge technology and human interaction. My interests span machine learning, natural language processing, and full-stack development, and this portfolio showcases some of the projects and research I’ve worked on along the way.

Projects

×
Health Insurance Application on Internet Computer Blockchain

Built a Health Insurance application on top of the Internet Computer Blockchain that automates the insurance claim settlement process using Canisters. Developed a priority algorithm with Convex Hull Trick and Square Root Decomposition, achieving O(√N log N) complexity in insertion, deletion and querying and reducing latency by 70x for 10⁵ inputs. Implemented a user-friendly interface using React that seamlessly interacts with the blockchain through Node.js.
×
Tamil Troll Meme Detection

I published this paper in ICON 2024 (21st International Conference on Natural Language Processing), where I applied machine learning and deep learning models to detect trolling in Tamil memes—combining both text and image features to address the problem in a low-resource language. By creating a labeled dataset and benchmarking multiple models, my work demonstrates how ML can distinguish between humor and offensive content in digital memes.
×
Project Winry

ProjectWinry is a Python-based personal voice assistant I built to explore natural language interaction, voice recognition, and simple automation. The assistant can take spoken commands and perform tasks such as searching the web, opening applications, or retrieving information, acting as a lightweight desktop companion.

Publications

TRO(F)LL or ROFL? : Exploring Troll Detection in Tamil Memes

Published at ICON 2024 (21st International Conference on Natural Language Processing), this paper applies machine learning and deep learning techniques to detect trolling in Tamil memes. Combining text and image analysis, the research tackles this low-resource language challenge by creating a labeled dataset and benchmarking various models. It demonstrates how ML can distinguish humor from offensive content in digital memes.

Travels

Documenting my travels around the world.

A video of the sun setting over the hills in the komodo islands, bali

Hey, I’m Aditya 👋

About Me

Projects

Health Insurance Application on Internet Computer Blockchain

Tamil Troll Meme Detection

Project Winry

Publications

TRO(F)LL or ROFL? : Exploring Troll Detection in Tamil Memes

Travels

Contact Me