• I’m a recent graduate from UC Berkeley with a Master’s degree in Analytics.

• I have a strong passion for data science and a proven track record of leveraging data to drive decision-making and improve business processes.

Interests:

• Machine Learning
• Deep Learning
• Data Science

Education:

• 🎓 Master of Analytics @ University of California, Berkeley
• 🎓 B.A. in Data Science @ University of California, Berkeley

Work Experience

Data Analyst @ Intex Recreation Corp. (Long Beach, CA | October 2024 - Present)

• Developed predictive models leveraging Prophet, Gradient Boosting, and Random Forest to forecast Intex airbed sales at Target, driving strategic decision-making and resource allocation.
• Analyzed weekly POS data to generate performance insights, identifying key drivers of sales fluctuations and informing actionable strategies.
• Automated reporting workflows for weekly POS data and monthly Supplier Business Plan (SBP) using Excel macros, improving process efficiency and accuracy.

Data Analyst Intern @ Synergies Intelligent Systems, Inc. (_Boston, NY | June 2024 - October 2024)

• Built a comprehensive dataset by extracting and preprocessing data from the COCO dataset and Open Images Dataset v7, focusing on images containing LED screens and various reference objects.
• Utilized YOLOv8 to accurately identify and locate LED screens and nearby reference objects within images, reducing object detection error by 13.3%.
• Leveraged the dimensions of reference objects to improve LED screen area estimation accuracy by 22%, applying homography and object detection techniques to achieve precise real-world measurements.

Data Analyst Intern @ Canadian Solar Inc. (Walnut Creek, CA | May 2023 - July 2023)

• Created a custom Excel macro, reducing report preparation time from 8 hours to 1 hour.
• Implemented model validation using Random Forest, XGBoost, etc., in Python to predict average inbound lead time with different storage locations, resulting in a reduction of forecasting errors from 23% to 8%.
• Deployed interactive Power BI dashboards for analysts to compare lead time, enhancing data-driven decision-making.

Data Engineer Intern @ Gotion Inc. (Fremont, CA | May 2022 – August 2022)

• Led the creation of Grafana dashboards for real-time battery data analysis, boosting operational efficiency by 18%.
• Built an AWS-based ETL pipeline processing 500,000+ data points daily, enhancing data management.
• Crafted optimized SQL queries for Grafana dashboards.
• Developed Python solutions to decode CAN (Controller Area Network), used in automotive electronics.

Data Analytic Intern @ RallyCry Ventures (Cambridge, MA | January 2021 – March 2021)

• Researched and uncovered early-stage startup companies that fit RallyCry’s investment criteria
• Used data operations and visualization tools like SQL and Excel to analyze real-world data on startups

Selected Projects

Airbnb Recommendation System

This project is an Airbnb Recommendation System designed to enhance user experience by providing personalized accommodation suggestions. By leveraging a combination of machine learning models—popularity-based, collaborative filtering, content-based, and hybrid approaches—the system accurately matches user preferences with available properties. The interactive dashboard, built with Dash and Plotly, allows users to dynamically adjust their preferences and receive immediate updates on recommendations.

Waste Classification

This project, the Automated Waste Classification System, utilizes advanced machine learning techniques to accurately classify different types of waste, such as paper, plastic, glass, and metal. This system aims to enhance waste management practices by promoting recycling and reducing landfill waste. By implementing models like Multilayer Perceptron (MLP), Vision Transformer (ViT), and Convolutional Neural Networks (CNN), we achieve high accuracy and reliability in waste classification. This innovative approach addresses critical environmental challenges, fostering a more sustainable future.

IMF PRGT Funding Classifier

The IMF PRGT Funding Classifier project aims to streamline the process of assessing eligibility for concessional loans under the Poverty Reduction and Growth Trust (PRGT) program. By leveraging advanced machine learning techniques, the project builds predictive models that analyze economic and financial indicators to determine which low-income countries are likely to qualify for zero-interest loans from the International Monetary Fund (IMF). This automated approach enhances efficiency, reduces manual review time, and ensures a more systematic evaluation process, ultimately supporting sustainable development and financial stability in eligible countries.