Developing Cybershuttle — a platform to connect your data and code to HPC, cloud, and local resources and run experiments anywhere.

Designing bots that navigate software and learn to perform complex tasks. Building RAG pipelines that use this learning to make software self-supporting.

Developed methods to diagnose ASD through EEG and IRT (w/IUPUC), diagnose ADHD through Eye Tracking (w/ODU), perform webcam-based eye tracking (w/UCSD), estimate cognitive load from real-time gaze data (w/ODU), and estimate webpage changes from archived crawl data (w/PSC). Worked as a TA for core CS courses, including Data Structures, Algorithms, Principles of Programming, and Computers in Society.

Explored interpretable methods to predict antimicrobial resistance from single-cell bacteria images. Trained Neural ODEs to approximate the dynamics of wave propagation in multimode fibers. Developed hypothesis testing methods to detect out-of-distribution inputs to deep neural networks.

Developed ML based solutions to extract URLs from PDF documents, archive their content for posterity, and generate robust links.

Developed a mobile app for video-based doctor consultations. Restructured cloud services on AWS to reduce costs, set up unified logging, and configured CI/CD. Co-led development for web and mobile apps.

Prepared course materials, conducted lectures, and graded student assignments for the CS2963 Presentation Skills course.

Created a knowledge base for WSO2 Support by scraping and vectorizing WSO2 docs. Created a document retrieval API for support questions, wrapped it with DialogFlow Conversation API, and integrated with Facebook Chat.