Projects

This project uses a YOLOv5 model trained on 369 images to automatically detect six classes of safety equipment on construction sites, ensuring compliance and worker safety. It streamlines monitoring processes, reduces manual inspections, minimizes reliance on safety officers, and demonstrates a commitment to safety and innovation, enhancing operational efficiency and company reputation.

Uses a Convolutional Neural Network (CNN) to classify concrete surface images as "defective" or "non-defective," improving inspection efficiency and accuracy. With 100 labeled images, sourced from airport construction projects, was preprocessed and split into 80% training and 20% testing. The CNN, with two convolutional layers and dropout for overfitting prevention, was trained over 20 epochs using the Adam optimizer and cross-entropy loss.

ECommerceU has several products that categorized based on sale in pareto. To increase the revenue, in this project analyze which products is the best fit to the others too bundling, and cross selling using apriori model analysis. Combine high-frequency product pairs with targeted promos, such as discounts or free shipping, to boost sales and revenue.

The conversion rate at E-CommerceU has decreased by 4.64% compared to the previous year. The analysis focuses on identifying the key factors using the Pareto principle and determining which funnel stage has the lowest conversion rate. The goal is identify causes and find improvement opportunities. Based on the findings, I will recommend ways to optimize the funnel and improve conversion rates for better sales.

E-CommerceU is facing challenges such as a decrease in GMV and user activity. To address this, I applied the RFM method to segment customers and identify key patterns. The goal is to improve marketing strategies for each group, boosting customer engagement and revenue. This approach creates personalized campaigns that better address each segment's needs.

E-CommerceU is a rapidly growing online platform focused on increasing revenue through data-driven strategies. As a data analyst, I developed a comprehensive dashboard to assist executive-level decision-making by providing actionable insights. This tool empowers leaders to identify trends, optimize operations, and drive business growth effectively.

Analyzed Iowa liquor sales performance to uncover sales trends, revenue generation, popular products, and regional sales performance. To provide clear insights, the analysis is visualized through an interactive dashboard, offering a comprehensive perspective on the data. These insights help identify opportunities for growth and optimize sales strategies.

Developed Tableau dashboard to provide a comprehensive helicopter view for C-Level executives, enabling data-driven decision-making. This dashboard provides actionable insights and a clear overview of the company’s performance, empowering C-Level executives to make strategic and well-informed decisions.

The investing company offers various investment products and aims to maximize profit within a limited campaign budget. To achieve this, I performed user segmentation and used logistic regression in Python to predict and mitigate churn rates. These insights help allocate resources effectively and enhance customer retention strategies.