Projects & Research
Exploring the intersection of machine learning, software engineering, and education through hands-on projects and research initiatives.
Leveraging Basis Alignment to create a Generalized Multi-Relational Graph Convolution Network in the Federated Setting
Developed a novel graph neural network architecture for knowledge graph completion in federated learning environments, incorporating differential privacy guarantees for secure distributed learning.
Imputing Parkinson's Patient Data with Machine Learning
Applied machine learning techniques to predict UPDRS scores for Parkinson's patients, using linear regression and clustering to analyze disease progression patterns in the Fox Insight dataset.
FPGAirPods
Designed and implemented real-time active noise cancellation on an FPGA using NLMS adaptive filtering, achieving low-latency audio processing for noise reduction.
RISC-V Processor
Implemented a high-performance RISC-V processor with two-way set-associative caches, achieving 1.5 GHz operation through optimized gate-level design and cache architecture.
Arduino Laser Tag
Built a location-aware laser tag system using ESP32 microcontrollers, integrating Google Geolocation API for real-time player tracking and game management.
Graph Neural Networks for Scene Segmentation
Investigated the application of GNNs to scene segmentation, developing novel methods for converting images to graph structures and implementing message-passing networks for pixel-wise labeling.
Uberprivate: Differentially Private Ride-Sharing
Designed a privacy-preserving ridesharing system using differential privacy and geo-indistinguishability, enabling users to control their location privacy while maintaining service utility.
Goodbye Dockerfile, Hello Bazel
Optimized CI/CD pipeline by migrating from Docker to Bazel, achieving significant performance improvements in build times and resource utilization.
Robotic Arm Rock Skipping
Developed a robotic system to analyze and optimize rock skipping in simulation, investigating the physics of stone skipping through precise control of release parameters.
LyRec: Lyric-Based Music Recommendation
Created a music recommendation system using song lyrics and sentiment analysis, implementing both TF-IDF and GloVe-based approaches for personalized recommendations.
Amazon Tenmarks Research
Conducted research on educational data analysis at Amazon Tenmarks, developing machine learning models to improve student learning outcomes and educational content recommendations.
WaferSat
Conducted thermal code testing for the WaferSat satellite-on-a-chip as a Researcher at MIT Space Systems Laboratory, contributing to the development of next-generation space technology.