About Me
I recently graduated with a B.Sc. in Computer Science and Engineering from Daffodil International University, Dhaka. During my undergraduate studies, I worked as a researcher in the DIU NLP & ML Lab, where I contributed to multiple peer-reviewed publications spanning Natural Language Processing, Computer Vision, transfer learning, and explainable AI — including work on insect diversity classification, skin disease classification, fraudulent transaction detection, and medical image segmentation.
My research interests center on Natural Language Processing, Large Language Models, Computer Vision, and multimodal learning. I am currently a Ph.D. student and Graduate Teaching Assistant at Boise State University, where I continue pursuing research in these areas.
Education
Ph.D. in Computer Science
Boise State University, USA
Pursuing doctoral studies in Computer Science.
B.Sc. in Computer Science and Engineering
Daffodil International University, Dhaka, Bangladesh
Graduated with a performance-based scholarship; ranked 4th in batch.
Experience
Graduate Teaching Assistant
Boise State University, USA
Graduate Teaching Assistant, supporting undergraduate Computer Science courses alongside doctoral studies.
AI Agent Developer
StudyNet
Building a RAG-based LLM system using a FAISS vector database and the Django framework. The custom agent helps students find information related to higher education.
Research Assistant
DIU NLP & ML Lab
Collaborated with a professor from Deakin University, Australia, on detecting fraudulent activities in bKash online transactions using NLP-based techniques.
Trainer, Advanced ML & DL Bootcamp
DIU NLP & ML Lab
Trained more than 30 students in the fundamentals of machine learning and deep learning, assisted with implementing deep learning concepts in code, and supported their paper publications.
Selected Publications
Enhanced U-Net Architecture with VGG16 Backbone and Attention Mechanisms for Automated Nasal Sinus Disease Segmentation
In press — 8th International Conference on Recent Trends in Image Processing & Pattern Recognition (RTIP2R), 2025
Explainable AI for Skin Disease Classification Using Gradient-Weighted Class Activation Mapping and Transfer Learning in Digital Health to Identify Contours
DIGITAL HEALTH, 2025 (Q2). doi:10.1177/20552076251404523
A Transfer Learning Approach to Classify Insect Diversity Based on Explainable AI
Discover Life, 55(1), 1-15, 2025 (Q2) — Link
Transfer Learning Techniques for Efficient Deep Neural Networks in Real-Time Strawberry Freshness Evaluation
International Conference on Innovations in Biotechnology and Health Sciences, Springer Nature Singapore, 2025, pp. 87-101 — Link
Vehicle Speed Detection System Utilizing YOLOv8: Enhancing Road Safety and Traffic Management for Metropolitan Areas
arXiv preprint arXiv:2406.07710, 2024 — Link
A Deep Learning Approach to Detect Complete Safety Equipment for Construction Workers Based on YOLOv7
arXiv preprint arXiv:2406.07707, 2024 — Link
Performance Analysis of LSTM and Bi-LSTM Models with Different Optimizers in Bangla Sentiment Analysis
2024 15th International Conference on Computing Communication and Networking Technologies (ICCCNT), IEEE, 2024, pp. 1-7 — Link
Projects
LLM-Powered PDF Summarization Tool
Built an end-to-end NLP application using the Hugging Face API and Streamlit to automate document analysis and summarization. The tool processes large PDFs, extracts key insights, and enables semantic search with interactive Q&A — showcasing practical applications of retrieval-augmented generation and LLM-based information extraction.
Nasal Disease Segmentation
Built a web platform using a U-Net model with an attention mechanism architecture and deployed it on Hugging Face Spaces for medical use.
Honors & Awards
- 2025 Performance-based Scholarship, Daffodil International University & Ranked 4th in Batch
- 2023 First Runner-up, Data Science Olympiad — 2nd International Conference on Big Data and IoT
- 2023 1st Runner-up, Intra University Startup Competition — DIU Accelerator Cup
- 2023 12th Position, Intra University Contest — Unlock the Algorithm Contest
Skills
Programming
Python, C, C++, Java
ML Frameworks & Libraries
TensorFlow, PyTorch, Scikit-learn, Keras
Visualization & Data
Matplotlib, Seaborn, OpenCV, PIL, Pandas, NumPy, EDA, Feature Selection
Tools & Platforms
Windows, Linux, Mac, LaTeX, Microsoft Office, Google Suite