MY PROJECTS
Welcome to my Projects page! Here, you'll find a showcase of my work and achievements in business analytics.
Each project highlights my varied skills in data analysis, data visualization, predictive modeling, and decision-making, demonstrating my ability to solve complex problems and drive informed business decisions.
FraudWatch:
Using SQL for Financial Fraud Detection
In this project, we harness the power of SQL to build a fraud detection system for financial institutions. By analyzing transaction records, customer profiles, and merchant data, our system identifies fraudulent patterns, ensuring robust security measures and customer trust. Through advanced SQL queries like CTEs, window functions, aggregate functions, case functions, views, subqueries and joins we enhance accuracy, efficiency, and regulatory compliance in fraud detection, ushering in a new era of financial integrity
Data Driven Strategy for Airbnb Ecosystem Optimization
This project aims to leverage advanced data analytics and machine learning to forecast the future performance of Airbnb hosts. By focusing on key metrics such as Superhost status and occupancy rates, we provide actionable insights to enhance host success and guest satisfaction on the platform. Utilizing Gradient Boost models for precision, the analysis encompasses exploratory data assessments to predict revenue trends and identify the tangible benefits of Superhost designation. This enables hosts to strategize for maximum financial returns, ultimately optimizing the Airbnb ecosystem.
Crossroads Classic Analytics Challenge NCAA TICKET Purchase Activity Prediction
This project developed an ensemble model combining XGBoost, LSTM, and Random Forest algorithms to predict NCAA ticket purchases and activity types. Leveraging the strengths of each method, the model achieved an impressive 98.51% accuracy. By analyzing various factors influencing ticket sales and user activities, the model provided precise predictions. This success was validated by securing the 5th rank out of 51 teams on the Kaggle Leaderboard, showcasing the model's robustness and effectiveness in real-world applications.
Data Mining:
Firm Collapse prediction
Firm collapse prediction has been a subject of interest for almost a century and it still ranks high among the hottest topics in economics. The aim of predicting financial distress is to develop a predictive model that combines various econometric measures and allows one to foresee a financial condition of a firm. The purpose of bankruptcy prediction is to assess the financial condition of a company and its future perspectives within the context of longterm operation on the market. Our model secured the 2nd place amongst all our peers at Purdue's MSBAIM cohort in a kaggle competition.
LEGO HISTORY DASHBOARD
Developed an insightful LEGO data dashboard using Tableau, winning $1,000 and securing 1st place in Purdue's "Love Data Visualization" competition, and represented Purdue at the Big 10 Love Data competition. This interactive dashboard explores LEGO sets, themes, and parts over 90 years, sourced from Kaggle. It showcases intricate data interactions and trends, visualizing changes in design and complexity, and tracking the diversification of themes. The project demonstrated advanced data visualization skills, presenting complex information in an accessible and engaging manner for enthusiasts and analysts.
This project identified Key Opinion Leaders (KOLs) for a new neurology product launch in the US. Utilizing data from clinical trials, research publications, leadership roles, and open payment records, we pinpointed influential experts. The process included prioritizing datasets, developing a methodology, and creating KOL profiles. The objective was to leverage these KOLs for product endorsement and promotion within the medical community, enhancing the product's credibility and reach.
TindeR x Netflix Integration
We aimed to integrate Tinder with Netflix datasets to enhance user engagement through personalized content discovery. This project allowed users to connect through shared Netflix interests, enhancing both social and entertainment value. Rolled out in two phases, we developed a Proof of Concept (PoC) with a Tableau dashboard and then then future scope to fully integrate the feature within Tinder. Using MS Project and Smartsheet, we effectively managed, scheduled, and monitored each phase, ensuring successful implementation and possibly revolutionizing user interactions on Tinder.
Leveraging LLMs in Medical Documentation Automation
In this project, we aimed to automate the medical documentation process, allowing healthcare providers to focus more on patient care. Using pre-trained LLMs, we created chains of reasoning to answer questions based on simulated doctor-patient conversations or dictations. This was part of the Purdue Data4Good Competition, highlighting our skills in data analysis, AI, and communication. The project showcased the potential of AI to enhance operational efficiency in healthcare by reducing the burden of paperwork.
This project is a data-driven exploration of global happiness and socio-economic factors using Tableau. It delves into the interplay between happiness metrics and socio-economic indicators, revealing nuanced patterns and correlations. Through intuitive Tableau visualizations, the complexities of well-being are unraveled, offering a compelling narrative that transcends borders. The project sheds light on the intricate tapestry of factors influencing happiness worldwide, providing deep insights into how socio-economic conditions impact global well-being.
Under CEO Art Peck, Gap eliminated the traditional Creative Director role, embracing a data-driven approach to fashion retail. This project examines the impact of this decision on Gap, Old Navy, and Banana Republic. Utilizing web-scraped data, Google Trends, and financial reports, the analysis reveals how data-driven strategies influenced design, marketing, and sales. The project provides insights into the effectiveness of this approach, highlighting shifts in consumer preferences and financial performance across the brands.
This project focused on designing a challenging yet fair final exam for an Operations Research course using SAS optimization techniques. By balancing difficulty and adhering to specific constraints, we aimed to create an exam that comprehensively tests students and prepares them for future success as data scientists. The project highlighted the adaptability and effectiveness of SAS optimization methods in educational settings, ensuring a rigorous assessment that meets educational standards.
Original Frozen Custard, a cherished family-owned restaurant in Lafayette since 1932, sought to modernize its operations using the power of SQL. This project aimed to centralize diverse data sources, streamline inventory management, and enhance decision-making. Through meticulous database design and implementation, the project optimized processes, identified star-selling products, measured employee performance, and elevated inventory management efficiency. The SQL-driven solution not only preserved the restaurant's rich traditions but also ushered in a new era of data-driven excellence.
This paper provides an in-depth exploration of the video game industry, focusing on how major companies like PlayStation, Nintendo, and Xbox navigate market trends and technological changes. It discusses several strategic frameworks, including "Blue Ocean" strategy, competitive positioning, and disruptive innovation. The document assesses how these strategies influence the sustainability of business models in the fast-paced gaming sector. By integrating academic theories with real-world case studies and industry reports, the paper offers a comprehensive view of the strategic maneuvers used by leading gaming companies to maintain and enhance their market positions amidst growing global competition.
In this project, we delved into the concept of masstige branding and pricing, exploring how brands bridge the gap between mass-market and luxury offerings. Masstige brands offer accessible yet aspirational products, combining quality and exclusivity. We examined its origins, notable collaborations, and key pillars such as perceived value and product differentiation. Additionally, we explored pricing strategies like value-based and promotional pricing. Understanding masstige dynamics is crucial for brands, especially in emerging markets like India, where rising middle-class demographics and digital penetration offer significant opportunities.