Developed a deep learning-based speech emotion recognition system that analyzes audio recordings and classifies human emotions such as happiness, sadness, anger, and neutrality. Extracted MFCC features from speech signals and trained neural network models to identify emotional patterns from audio data using datasets such as RAVDESS and TESS.

Engineered a machine learning pipeline achieving 85% accuracy in predicting customer booking behaviour and developed a scalable demand forecasting framework supporting capacity planning across 1,500+ premium lounge departures at Heathrow Terminal 3.

Architected a deep learning model for recognizing handwritten digits and characters from images using Convolutional Neural Networks (CNNs). Trained on MNIST/EMNIST datasets with image preprocessing and classification techniques, with potential for extension to full word and sentence recognition using sequence models such as CRNNs.

Developed a quantitative pricing framework for coffee futures options by integrating Cost of Carry, Black-Scholes, and Monte Carlo simulation models to value commodity derivatives and analyze pricing under real-world market conditions.

Built an end-to-end machine learning pipeline to predict customer churn by engineering customer-level features, optimizing ensemble models, and using SHAP to deliver interpretable insights into the key drivers of customer attrition.

Developed a generative AI-powered productivity application using Amazon PartyRock to automate task execution, streamline workflows, and generate intelligent, context-aware responses through natural language interactions.