GPA: 3.9 / 4.0

Coursework: Analysis of Algorithms, Abstract Algebra I & II, Computational Learning Theory, Machine Learning, Deep Learning for Computer Vision, Probabilistic Models for AI, Real Analysis, Fourier Analysis, Topology, Quantum Mechanics I & II, Statistical Physics

Transferred to Columbia University

Coursework: Compressive Sensing & Signal Processing, Causal Inference, Topology, Honors Real Analysis, Number Theory, Operating Systems, Statistical Learning Theory, Machine Learning, Deep Learning, Relativity and Quantum Physics

GPA: 3.7 / 4.0

Coursework: Real Analysis, Abstract Algebra, Number Theory, Mathematical Statistics I & II, PDEs, ODEs, Proofs, Linear Algebra, Multivariable Calculus, Engineering Statistics, Software Development & Design, Discrete Structures, Software Components, Mechanics I & II

• Incoming Software Engineer Intern on the AWS Nitro team.

• Currently researching Graph Neural Networks for relational learning and representation learning on structured data.
• Designing and benchmarking experiments, evaluating model behavior, and iterating on architectures with Prof. P. Chin.

• Selected as one of 80 students out of 10,000 nationwide for an intensive program in mathematics and programming.
• Studied number theory and combinatorics in depth; collaborated on complex problem sets.
• Applied OOP principles, algorithms, data structures, and graph theory in weekly projects.

• Assisted teaching CSE 2231: Software Development and Design — OOP principles, design patterns, and software engineering best practices.
• Conducted lab sessions and office hours, supporting students with debugging and assignments.

• Built automated testing frameworks achieving 35% reduction in manual testing and 20% increase in bug detection.
• Created custom automation solutions for data processing, increasing operational efficiency by 40%.

• Developed ML models to analyze travel data, leading to an 18% increase in customer retention and 22% boost in booking accuracy.
• Automated competitor benchmarking data aggregation, reducing analysis time by 40%.
• Integrated market intelligence into platform enhancements, improving UX scores by 25%.

• Built custom React components using Shadcn/UI and Radix UI to deliver reusable, product-specific UI building blocks.
• Implemented composable component APIs and interaction patterns to improve consistency, accessibility, and development speed across product surfaces.

• Advised 200+ clients monthly on VoIP, MPLS, and SIP solutions; drove 12% increase in contract renewals.
• Designed technical proposals achieving 25% rise in sales conversions.

Full-stack virtual clinical trial platform using a multi-agent AI pipeline to simulate, validate, and replicate randomized controlled trials with 3D visualization and OMOP CDM integration. Backend services can be hosted on AWS.

Research intelligence tool that ingests biomedical PDFs and generates interactive force-directed knowledge graphs, with RAG-based retrieval and Claude-powered hypothesis generation. Backend APIs can be hosted on AWS.

Financial forecasting model trained on 2007–2018 data, tested on 2019–2022. Achieved a Sharpe ratio of 1.1 using Pandas, Scikit-learn, and TensorFlow, with backend inference services deployable on AWS.

Real-time emotion detection system with on-device ML inference. Includes assistive speech detection for individuals with speaking disabilities.

IoT-enabled automatic pill dispenser built with Arduino and Raspberry Pi. Features scheduling, reminders, encrypted communication, and a mobile monitoring app.

Computational framework analyzing NC congressional redistricting fairness using SKATER clustering for community identification and ReCom MCMC ensemble analysis via GerryChain.

Data-driven market segmentation for commercial real estate using K-means clustering, Silhouette analysis, and PCA to identify five distinct property segments based on occupancy patterns and resilience metrics.

Movie recommendation system using matrix factorization and neural collaborative filtering. Built with PyTorch and a from-scratch NumPy implementation, with an interactive Streamlit web app.

Differential sensing system using Python and OpenCV for anomaly detection in real-time data streams, achieving 95% accuracy with a Plotly dashboard.

Data analysis of Divvy bike share usage through cleaning, exploratory analysis, and visualization to uncover patterns in ride types, user categories, and popular stations.