– Led multiple research projects including Fake Job Classification, Sentiment Analysis of Covid Vaccine Opinions, and ECG Classification on PhysioNet.– Scraped and labeled 10,000+ tweets for sentiment analysis of the Pakistani public regarding Covid vaccines, fine-tuned BERT, and achieved a 95.62 F1 score. Analysis revealed 42% neutral, 41% positive, and 17% negative sentiments.– Designed a hybrid residual-inception network for ECG classification on a Kaggle PhysioNet dataset, achieving a third-best score of 50.87%.

– Collaborated with cross-functional teams to develop end-to-end data-driven solutions for NFT price prediction and multimodal in-bed pose estimation. Performed preprocessing, feature engineering, and analysis on different datasets.– Implemented automated data collection pipelines, including web scraping and API integrations for diverse data sources like OpenSea, social media platforms, integrating crypto and commodity prices from Binance and Yahoo Finance.– Created annotation tools on Tkinter and interactive web applications for Decentraland and Sandbox, presenting micro-feature information. Contributed to model development, achieving 68% accuracy in price prediction and 71% PCKh@0.5 performance in pose estimation.

Provided machine learning and application development services. Completed projects such as: Automated Test Checker and Feedback Generator, Resume Ranking System, and Soccer Score.

Assisted with planning the syllabus, organizing guest lectures, designing course and lab assignments, and collecting feedback for CS469: Data Science

Built an image recovery application for stolen phones and worked on bypassing Android smartphone passwords.