Mahmud Hasan

I'm a final-year Computer Science undergraduate with hands-on experience in machine learning, deep learning, and healthcare AI. I build analytical and AI based systems using real-world datasets, with a strong interest in applied machine learning, healthcare analytics, and scalable data solutions.

My thesis, DS-GCA, developed a multimodal deep learning framework for breast cancer survival prediction fusing histopathological whole-slide images with structured clinical tabular data. I fine-tuned a Swin Transformer and used Attention-Based Multiple Instance Learning (ABMIL) with a Dual-Stream Gated Cross-Attention fusion module to combine the two modalities, validated through 5-fold cross validation.

I also designed and experimentally validated a carbon aware AI inference framework a Green AI approach that dynamically switches between pruned models based on real-time grid carbon intensity, reducing inference-phase emissions while maintaining accuracy.

Beyond research, I build full stack applications using the MERN stack and PHP/MySQL, focused on real-world usability and robust engineering.

I'm looking to apply this research to deployment experience in an applied ML or AI research engineering role, particularly in healthcare AI or sustainability focused AI.