I'm a developer focused on agent-powered workflows, full-stack web apps, and the messy little problems most people learn to live with.
I'm a Computer Science junior at UConn who likes building tech that actually makes it out of the IDE and does something useful in the real world.
My focus is on API development, full-stack apps, and emerging tech — and I learn best by shipping things, breaking them, and shipping them again.
Long term, I want to be a well-rounded engineer who builds efficient systems that deliver real value, and keeps growing alongside the field.
An AI-powered job application tracker that processes natural language commands and automates multi-step application workflows through an agent-based architecture. Backed by the Anthropic Claude API and synced to Google Sheets in real time for persistent tracking and analytics.
A mobile app that rethinks budget travel across Asia by surfacing multi-city stopover routes that often cost less than direct flights — while letting travelers explore more destinations along the way. Pulls real-time fares from the Amadeus API and ranks itineraries by cost, layover convenience, and total travel time.
An automated ETL pipeline that ingests, transforms, and loads sales data into PostgreSQL on Amazon RDS using AWS Glue and Python — orchestrated end-to-end with Jenkins CI/CD for consistent, repeatable runs. Cloud infrastructure (RDS, S3, IAM) is provisioned with Terraform and CloudFormation, and the pipeline auto-generates Excel reports via OpenPyXL to make the transformed data easy to analyze.
A starter QA automation framework that brings API smoke tests and UI flows under one Maven build and one Jenkinsfile. RestAssured covers HTTP endpoints, Selenium WebDriver handles browser flows, and the included Pipeline-as-Code setup makes the whole suite CI-runnable from day one.
A computer vision system that reads emotions off a live webcam feed in real time. Built with Python and OpenCV for video capture and processing, with DeepFace classifying expressions into 3+ categories (happy, sad, neutral, and more). Integrated MTCNN for accurate face localization, which dramatically improves detection reliability in low-light and noisy conditions.
Open to internships, full-time roles, side projects, and the occasional good conversation.
Hiu3648@gmail.com