Won 2nd place among 120 students in the CSE 472 Machine Learning Lab competition at BUET, developing a custom CNN architecture for Bangla digits recognition implemented from scratch using only NumPy and OpenCV. Achieved 95.9% test accuracy on the NumtaDB benchmark by implementing every convolution, pooling, activation, and optimization step manually.
View Project Details

Selected as one of the 40 finalists among numerous teams from across the country with our project "Blockchain Based Ticketing Platform". Participated as part of the "Recursively Enumerable" team with Ataf Fazledin Ahamed and Md. Tanzim Azad Nishan.

For demonstrating outstanding performance by completing numerous tasks and resolving critical issues, supporting team processes, and developing essential features that significantly improved the product.
IQVIA Impact Program – Silver (Certificate)