Subramanian
Mathur Seetharaman

Developed an interactive Power BI dashboard analysing 738K+ UK flights (Jan to May 2025) using Civil Aviation Authority (CAA) data to uncover delay patterns, airline punctuality, and airport operational efficiency.

• Overall on-time performance stood at 68.82%, with an average delay of 12.13 minutes, severe delays at just 4.20%, and a cancellation rate of only 0.01%
• Heathrow contributed the highest share of total delay minutes (29.98%) across six major UK airports, while Lufthansa led airline punctuality with 74% on-time performance
• March recorded the lowest average delay, with delays peaking in January and May, revealing clear seasonal variation in flight performance

Compared K-means, DBSCAN, GMM, and Hierarchical clustering algorithms on synthetic high-dimensional data and the UCI Human Activity Recognition dataset.

• Evaluated performance using external (ARI) and internal (Silhouette) metrics
• Analysed the impact of dimensionality reduction using PCA
• Assessed algorithm resilience to noise and irregular cluster shapes

Independently provisioned and configured a complete Microsoft Dynamics 365 Finance & Operations local development environment.

• Navigated Microsoft Lifecycle Services (LCS) to provision resources
• Downloaded and deployed the VHD (Virtual Hard Disk) for local VM staging
• Configured the architecture for subsequent SQL and Power BI reporting

Analyzed 160,000+ e-commerce transaction records to identify customer behaviour patterns, sales trends, and product performance.

• Performed customer segmentation using K-Means clustering
• Built Random Forest models to analyze product conversion (R² = 0.87)
• Modelled daily sales trends to discover seasonal promotion effects

Investigated dog rehoming durations across specific breeds (Bichon Frise, Border Collie, Dobermann) to identify adoption trends.

• Performed comprehensive data cleaning and exploratory analysis
• Fitted statistical distributions and estimated confidence intervals
• Conducted pairwise comparisons to evaluate breed-specific differences