← Back to Home

Other Academic Projects

A quick snapshot of academic work across strategy, risk, databases, and responsible AI-each project summarized with the core objective, approach, and key outcomes.

Under Armour Strategic Transformation project cover image

Under Armour-Strategic Transformation & Analytics Vision

Strategy Analytics Vision Product + Ecosystem

Built a transformation plan centered on three pillars-personalized performance insights, sustainable manufacturing (DfM), and smart training ecosystems-to evolve Under Armour into a data-driven athlete platform.

  • Defined a roadmap from data foundation → sustainable manufacturing shift → smart performance launch → global athlete network expansion.
  • Identified key risks (data integration complexity, supply volatility, adoption of smart tech) and mitigation via phased rollout, diversified sourcing, and athlete feedback cohorts.
ERM for Vighnaharta Food project cover image

From Risk to Resilience-ERM for Vighnaharta Food (Dairy Manufacturing)

Enterprise Risk Management Operations Risk Framework

Designed a practical ERM framework for a small dairy manufacturer, moving from reactive firefighting to structured risk governance, measurement, and mitigation planning.

  • Created risk buckets (supply chain, quality control, financial, market, sustainability, regulatory, technology) and a quantitative scoring approach to prioritize threats.
  • Proposed lightweight governance: risk champion + shared risk log, short “risk moments” in meetings, and simple visual tracking for early warning signals.
Petal Post SQL project cover image

Petal Post-A Plant Delivery System (SQL Database Project)

SQL Database Design E-commerce System

Designed a plant e-commerce concept combining a web front-end (Power Pages) with an SQL-backed system to manage customers, products, and orders for a smooth plant-buying experience.

  • Defined an end-to-end flow from browsing plants to checkout, supported by a structured backend in SQL/Azure Data Studio.
  • Focused on reliable data storage and retrieval for core entities (plants, orders, customers) to keep the shopping journey fast and consistent.
RoyaltEase system project cover image

RoyaltEase-Digital Rights & Royalty Management System

Systems Design DRM + Royalties Data Flows

Proposed a platform to streamline digital rights and royalty management with automated royalty calculation, usage tracking, and secure transaction handling.

  • Mapped the system using context + DFD levels, clarifying how artists, users, monitoring, rights rules, and payments interact.
  • Modeled key entities and relationships (artists, content, usage, license, royalty, payment, monitoring logs) to support traceable, auditable royalty distribution.
Prediction of Energy Consumption project cover image

Prediction of Energy Consumption

Machine Learning Regression Feature Engineering

Built predictive models to estimate household energy consumption using building characteristics and weather variables, translating raw data into planning-oriented insights.

  • Performed targeted cleaning and feature creation (e.g., filtering comparable homes and engineering additional variables for stronger signal).
  • Compared models like CART (classification/regression trees) and linear regression to understand drivers and predictive performance.
Netflix vs Prime EDA project cover image

Netflix vs Prime – User Behavior Analytics

Python Pandas Exploratory Data Analysis

Comparative analysis of Netflix and Amazon Prime user datasets to understand demographic patterns, subscription behavior, and device usage trends across platforms.

  • Cleaned and structured two independent streaming datasets for consistent comparison.
  • Analyzed age distribution, gender breakdown, subscription models, and device engagement patterns.
  • Restructured the project into a reproducible analytics workflow separating notebooks from production-ready scripts.
The Mind-Reading Illusion project cover image

The Mind-Reading Illusion-Why Social Media Shows You What You Never Search or Say

Responsible AI Recommender Systems Privacy + Ethics

Investigated why feeds can feel “psychic” by testing how micro-signals (dwell time, pauses, replays, social-graph effects) reshape recommendations even without explicit searches or typing.

  • Ran user-side interaction experiments (micro-pause, social graph influence) to observe how quickly recommendation outputs shift.
  • Reviewed platform documentation/model-card style sources to compare what platforms disclose vs. what users assume—framing the gap as an ethics and transparency issue.