Projects

Led a data science project for DoorDash predicting delivery durations using a dataset of 200,000+ orders and 20+ columns. Utilized Python (Pandas, NumPy, Scikit-learn, Matplotlib, Seaborn, XGBoost) for data preprocessing, feature engineering, and modeling. Implemented an optimized XGBoost model achieving an average deviation of 13.87 minutes from actual delivery times, explaining 40% of variability. Uncovered insights: peak hours (7 PM - 4 AM) account for 90.4% of orders, with 177.2% higher weekend volume. Identified 6 markets with high peak-to-off-peak delivery time ratios (>75th percentile) and 6 with low ratios (<25th percentile). Created visualizations including correlation heatmaps, order density plots, and time series analysis. Findings informed market-specific strategies for staffing, pricing, and operational improvements, enhancing DoorDash's delivery efficiency.

Spearheaded the development of CampusCircle, an all-in-one platform designed to enhance campus life for over 600 UCI organizations. Built a system connecting students with club events through real-time updates, an integrated Google Map, and gamified experiences. Implemented a points system where students earn points for attending events and clubs earn points for student participation. Seamlessly translated Figma designs into functional code, integrated interactive map pins, and developed an RSVP system for event creation and participation. Ensured efficient data management with intuitive UI/UX, creating a more connected and thriving campus environment.

UCI Datathon 2024 1st Place: Best Analysis of Airbnb Market in Dublin [Sponsored by StrataScratch]. Given StrataScratch datasets, our team performed an Exploratory Data Analysis (EDA) to analyze, visualize, and present market demand and supply, aiming to elevate the Airbnb experience for guests and hosts. Leveraged tools such as Deepnote, Alteryx Designer, and Python libraries (Pandas, Matplotlib, Seaborn, NumPy, Scikit-Learn). Built visualizations such as a Day & Month Performance Heatmap, Geographical Heatmap, and tested/trained a Random Forest Classifier machine learning model to achieve 89.35% accuracy.

36-hour Hackathon. Revolutionizing food sustainability with a web app that empowers users to manage ingredients, expiration dates, and access custom recipes with an email login system.

Freelance Photographer with 4+ years of expertise in capturing compelling moments. Specializing in portrait, event, and automotive photography, I am dedicated to creating stunning visuals that leave a lasting impression. Explore my world of captivating imagery by clicking the link. Let's create timeless memories together.