Kaniz Samiha Tashfi
Founder & Business Operator · Browmein
BRAC University
Project
Developed a multi-view deep learning model combining tabular, sequential, and graphical representations using attention-based fusion. Achieved 98.27% accuracy, 19.1% lower inference latency, 75.99% lower computational complexity, and 4.1× smaller model si
Project
Developed a journaling file system with transaction logging and crash recovery. Implemented block allocation, inode/data-block management, free-space tracking, directories, and POSIX disk I/O.
Project
Developed a full-stack laundry booking platform with service management, reviews, ratings, feedback, and appointments. Collaborated in a 4-member team and deployed the application on Vercel.
Project
Built a machine learning model to classify software quality using complexity, churn, bugs, testing, and developer-experience features. Preprocessed 1,600 records and compared multiple classification algorithms using standard evaluation metrics.