Junior at the University of Georgia studying Computer Science, working on projects and gaining real-world experience while finishing my degree.
A selection of class projects and practical work that show the languages, tools, and problem solving I use while completing my degree.
Building an AI wardrobe app that auto-tags clothing by category, color, and style via CLIP zero-shot Designing “find similar pieces” visual search over image embeddings in pgvector with HNSW indexing Developing an outfit recommender ranked by occasion and live weather; launching to UGA fashion orgs
Built a full-stack simulator of airline crew recovery after an ATL ground stop (186 flights, 61 crews) Designed an hourly OR-Tools CP-SAT optimizer that cut flight cancellations 44% vs. no reassignment Trained an XGBoost delay model (AUC 0.67 vs. 0.60 baseline); served via FastAPI to a React dashboard
Junior at the University of Georgia studying Computer Science, working on projects and gaining real-world experience while finishing my degree.
I am Ash Sun, a junior at the University of Georgia studying Computer Science. I have some professional experience and am building toward a full-time career in software. I focus on learning practical skills and shipping projects while finishing my bachelor's degree.
Experience
Application Development Intern
Fiserv · Jun 2026 – Aug 2026
Software Engineer Intern
AI @ University of Georgia · Mar 2025 – Sep 2025
ML Data Engineer Intern
Prithvy (Environmental Research Lab) · Jun 2025 – Aug 2025
Junior Engineer Intern
Upgrade AI · Jan 2025 – Mar 2025
Education
The University of Georgia
B.S. in Computer Science, Business Minor; GPA: 3.4/4.00 · Class of 2028
Skills
Building an AI wardrobe app that auto-tags clothing by category, color, and style via CLIP zero-shot Designing “find similar pieces” visual search over image embeddings in pgvector with HNSW indexing Developing an outfit recommender ranked by occasion and live weather; launching to UGA fashion orgs
Built a full-stack simulator of airline crew recovery after an ATL ground stop (186 flights, 61 crews) Designed an hourly OR-Tools CP-SAT optimizer that cut flight cancellations 44% vs. no reassignment Trained an XGBoost delay model (AUC 0.67 vs. 0.60 baseline); served via FastAPI to a React dashboard