Aakash Mehta

Key Accomplishment: Built dashboards and reporting tools that helped teams better understand cloud usage, reduce manual reporting work, and make clearer data-driven decisions. • Built enterprise Power BI dashboards analyzing Snowflake usage and cost patterns, improving cost visibility and driving a 15% reduction in cloud spend. • Replaced manual reporting with leadership dashboards using Power BI and DAX, reducing data validation effort by 80%. • Defined KPI frameworks and optimized reporting workflows, improving data quality checks and reducing recurring validation time by 30%. • Collaborated with cross-functional teams to deliver enterprise analytics solutions adopted across multiple business units. • Supported a company-wide GenAI initiative (~200 employees) by identifying operational pain points and promoting AI tool adoption through structured innovation frameworks.

Key Accomplishment: Built a course recommendation system using learning data from multiple platforms to help suggest more relevant courses and improve how users discover content. • Built a data-driven course recommendation system by integrating 50,000+ learning records from Udemy, LinkedIn Learning, and MasterClass into a unified dataset. • Improved recommendation accuracy by 70% using collaborative and content-based filtering techniques. • Increased click-through rate (CTR) by 40% by improving personalization and ranking of recommended courses. • Reduced bounce rate by 20% by better matching recommendations with user interests and learning behavior. • Conducted exploratory data analysis and model validation using Python (Pandas, NumPy, Matplotlib) to evaluate and improve recommendation performance.

During my Master’s in Data Analytics Engineering at Northeastern University, I completed applied projects across healthcare, environmental forecasting, NLP, and labor economics. I analyzed U.S. holiday travel and air quality data, identifying a 370% PM2.5 spike and improving travel demand forecasting accuracy by 40% using Power BI. In healthcare analytics, I built Logistic Regression and Random Forest models on data from 130 U.S. hospitals, achieving 88% accuracy in predicting diabetic readmissions. I also classified 500,000 song lyrics using TF-IDF and XGBoost with 75.4% accuracy, and used SQL/NoSQL to analyze wage disparities, reducing insight generation time by 40% and improving efficiency by 60%.

During my internship at Shah and Anchor Kutchhi Engineering College, I developed and deployed an Android-based Course Recommendation System using collaborative and content-based filtering techniques. I integrated 50,000+ learning data points to improve recommendation accuracy by 70%, increase click-through rate by 40%, and reduce bounce rate by 20% through personalized modeling and feature optimization. Additionally, I designed and built an independent blockchain-based project focused on cryptocurrency fundamentals, implementing core functionalities using the Swagger library and RESTful APIs to simulate decentralized transaction workflows.