CSE graduate and aspiring AI Engineer building localized NLP solutions for Bangla language and education. IEEE-published researcher with hands-on experience in deep learning, LLMs, and full-stack AI systems.
I'm a Computer Science and Engineering graduate from Daffodil International University, passionate about building AI systems that solve real problems for underserved language communities — particularly Bangla.
My final-year project focused on automated MCQ generation from Bengali text using large language models, and my IEEE-published research explores gender classification in Bangla NLP. I believe impactful AI starts with solving local, specific challenges.
Beyond research, I've been active in the tech community — as an IEEE student branch executive, ICPC volunteer, and competitive programmer with an ICPC Honorable Mention.
Designed and built an end-to-end AI system to generate Bengali multiple-choice questions with distractors and correct answers from text passages. Evaluated BanglaT5, Gemma, Ministral, and Qwen3-8B — selecting Qwen3-8B based on BLEU, METEOR, ROUGE-L, and BERT-F1 scores. Includes a Streamlit interface for classroom use and a custom Bengali MCQ dataset sourced from NCTB books, GK books, and Wikipedia.
Researched and developed a deep learning pipeline for gender classification from Bengali text, comparing multiple architectures for accuracy on low-resource Bangla data. Paper accepted and virtually presented at the 16th IEEE International Conference on Computing, Communication and Networking Technologies (ICCCNT 2025) at IIT Indore, India.
Open to research collaborations, AI/ML internships, grad school opportunities, and conversations about Bangla NLP.