Researcher in Artificial Intelligence
Exploring the frontiers of Large Language Models, Artificial General Intelligence, and Deep Learning. Currently pursuing BSC in Computer Science with focus on next-generation AI systems.
Selected publications and ongoing investigations
An eight-layer retrieval-augmented pipeline that answers Bengali and Banglish student questions from NCTB textbooks, with 95% subject detection accuracy and a 0.0 hallucination rate on formal queries.
Read MoreTwo-tier deep learning framework for urban traffic monitoring using acoustic horn analysis. Published at IEEE QPAIN conference.
View DOIDeep learning framework for stenosis detection in low-resource settings, addressing critical healthcare diagnostics through computer vision.
Read MoreUser-independent deep learning approach for surface electromyography signals, enabling intuitive assistive technology interfaces.
Read MoreCurrent investigations and theoretical interests
Investigating emergent capabilities, fine-tuning methodologies, and alignment techniques in foundation models. Focus on efficient adaptation and domain-specific applications.
Exploring architectures for general-purpose reasoning, multi-modal understanding, and autonomous agent systems with robust safety considerations.
Neural architecture search, optimization dynamics, and interpretability methods. Emphasis on efficient training and deployment in resource-constrained environments.
Medical imaging analysis, real-time detection systems, and multi-modal perception. Application of vision transformers and diffusion models.
Latest activities and announcements
AdaLLM, a subject-adaptive RAG framework for NCTB Class-8 Science & BGS question answering, accepted at the 2nd International Conference on Advances in Computing, Communication, Electrical, and Smart Systems.
HornNet, a two-tier deep learning framework for real-time vehicle classification from acoustic horn signatures, published and presented at the 2026 IEEE 2nd International Conference on Quantum Photonics, Artificial Intelligence & Networking.
Two papers on coronary artery disease detection and sEMG-based wheelchair control accepted at International Conference on Big Data, IoT and Machine Learning.
Launched personal academic website to share research progress, publications, and technical implementations in AI/ML.
Interested in collaboration, research discussions, or academic opportunities? Feel free to reach out.