A web-based dashboard for visualizing statistical distributions, hypothesis tests, and regression outputs. Built to help students understand core stat concepts visually.

This very website — an open-source, data-driven Next.js site for the Statistics Programming Club. All content is managed through JSON files with a custom visual editor.

A curated, well-documented collection of Jupyter notebooks covering core ML algorithms from scratch — linear regression to transformers — maintained for club workshops.

A smart attendance management system for university departments, featuring QR-based check-in, automated reports, and department-level analytics.

An IoT-integrated smart farming tool that uses sensor data and ML models to predict crop health, water needs, and yield estimates. Winner of DU Tech Fest Hackathon 2024.

Research project on natural language processing for Bangla — a low-resource language. Focuses on named entity recognition and sentiment analysis using fine-tuned transformer models.

Developing a hybrid algorithm that utilizes quantum annealing for hard combinatorial problems while leveraging classical gradient descent for fine-tuning.

Automating the design of efficient deep learning architectures optimized for low-power mobile and IoT hardware without sacrificing accuracy.

Establishing a mathematical framework for fairness and accountability in AI systems, aimed at creating global standards for algorithmic transparency.