My Projects

A collection of my academic and professional projects showcasing work in data engineering, analytics, database design, and web development. Each project reflects my ability to solve real-world problems using data-driven approaches and modern technologies.

Academic Research Repository (Database Design Project)

Designed and developed a relational database system to manage academic publications, research projects, and grants. The repository supports researchers, students, and administrators with features like publication tracking, collaboration tools, citation analysis, and grant management. Implemented an ERD with 12 normalized tables, enforced business rules, and optimized queries for efficient data retrieval.

Multi-Disease Prediction Using Deep Learning

Database Systems & Web Application Projects

BrewShop: Online Coffee Marketplace

Irvington Historical Society Web App

Data Engineering & Big Data Processing Projects

Instacart Market Basket Analysis

Retail ETL Pipeline (Colab + PySpark)

Building Footprint Extraction (Deep Learning)

Data Visualization & Business Intelligence Projects

New York Housing Market Analysis (R)

New York Airbnb Dashboard (Dash/Plotly)

NBA 2024–25 Dashboard (Power BI)

NYC Housing Dashboard (Looker Studio)

Credit Risk & Default Analysis – Home Credit Dashboard (Tableau)

Machine Learning & Predictive Modeling Projects

Smart Vehicle System – Real-time hazard detection system

Statistical analysis of Instacart data - Exploratory Data Analysis

This three-tier e-commerce platform was built to streamline the purchase and management of coffee products. I integrated features like reviews, coupons, wishlists, and automated invoice generation with TCPDF. On the admin side, the system included tools for inventory, payments, shipping, and customer support. The result was a user-friendly application combining smooth workflows with robust database-driven logic.

Note: No access to the website or source files. This was a collaborative academic project, and sharing of files is restricted per project guidelines.

I developed a digital archive and mapping system to preserve the neighborhood house histories of Irvington. Using Leaflet.js and OpenStreetMap, the app displayed interactive maps with clickable house markers linked to historical data. A geocoding workflow was added to handle address lookups and predictive search. The platform also included inline editing and archival-safe deletion of records to maintain historical integrity.

Note: Access to project materials is temporarily restricted pending stakeholder permission. Once approved, detailed documentation and visuals will be published.

This project focused on statistical analysis of the NYC housing market using R. I explored neighborhood-level pricing, availability patterns, and long-term market trends. Predictive models were built to forecast housing demand and property values. The findings provided valuable insights into real estate behavior and supported data-driven investment decisions.

I designed and published a housing market dashboard in Looker Studio for my portfolio. The dashboard presented neighborhood-level metrics, pricing trends, and property availability. Multiple datasets were combined to give a holistic view of real estate patterns. The visuals were optimized for sharing on LinkedIn, highlighting my professional branding and BI skills.

This dashboard was designed to track player and team performance across the 2024–25 NBA season. It featured visuals for shooting accuracy, efficiency metrics, and key performance indicators. Interactive slicers enabled filtering by player, team, or game date. The project served as a portfolio piece showcasing my Power BI and DAX expertise.

I created an interactive dashboard with Plotly and Dash to analyze New York City’s Airbnb dataset. The dashboard allowed exploration of occupancy rates, pricing trends, and host activity across neighborhoods. Filters and visualizations were designed for intuitive comparison between regions. This project demonstrated my ability to turn large datasets into accessible visual stories.

This project involved processing over 3 million transactions using PySpark and SparkSQL. I analyzed customer behavior to predict product reorders and segment buyers into categories. Demand forecasting was performed to support smarter inventory planning. The workflow demonstrated my ability to scale analytics with big data tools.

Using satellite imagery, I applied ResNet and SkyNet models to extract building footprints for urban planning. The model enhanced detection accuracy and contributed to geospatial analysis applications. This approach supported better infrastructure mapping and smarter city planning. It demonstrated how deep learning can be applied to real-world environmental datasets.

I built an end-to-end ETL pipeline for retail sales data using PySpark in Google Colab. The pipeline ingested raw datasets, transformed them into structured outputs, and prepared them for analytics. Automation was included to streamline data preparation and reporting. This project highlighted my skills in data engineering and reproducible workflows.

This project applied Random Forest, Naïve Bayes, and Decision Tree algorithms to predict diseases from medical datasets. The system evaluated model accuracy and efficiency across multiple conditions. It aimed to support early detection and medical decision-making. The project demonstrated the potential of machine learning in healthcare analytics.

I developed a smart vehicle assistance system to improve driver safety in hazardous conditions. Machine learning models analyzed sensor data to detect risks and provide alerts. The project focused on real-time prediction of road hazards. It illustrated how AI can enhance transportation safety and driver awareness.

Note: Lost access to original project files and datasets. The overview provided reflects the project’s key objectives, implementation, and outcomes.

Developed an interactive Tableau dashboard to analyze customer credit behavior and identify potential default risks. Integrated data on income, loan amount, repayment history, and late payments to uncover risk patterns. Built KPI cards, risk segmentation visuals, and correlation heatmaps for decision support. Enabled lenders to assess borrower reliability through data-driven insights and predictive indicators.

Performed exploratory data analysis on millions of Instacart grocery transactions to uncover shopping patterns and customer behavior. Analyzed order frequency, product reordering trends, and time-based purchasing habits. Visualized insights using Python (Matplotlib, Seaborn, Pandas) for trend identification and correlation mapping. The findings provided a foundation for predictive modeling and customer segmentation strategies.

Case Study on Research Papers

Conducted an in-depth analysis of the research paper “Perception! Immersion! Empowerment!” exploring how human cognitive superpowers can inspire effective data visualization design. The study summarized how visual perception, interactive immersion, and analytical empowerment enhance user understanding of data. Reflected on techniques that connect storytelling with visual cognition. The case study emphasized bridging design thinking with data communication principles.

Superpowers as Inspiration for Data Visualization

The Algorithms That Make Instacart Roll

Reviewed and summarized the paper “The Algorithms That Make Instacart Roll,” highlighting the data science processes that drive personalized grocery recommendations. Analyzed how machine learning models predict product reorders, optimize inventory, and improve user experience. Explored key algorithmic principles behind order prediction and recommendation systems. The case study demonstrated the intersection of big data analytics, customer behavior, and operational efficiency.