I'm Md. Nasif Safwan, a dedicated Computer Science & Engineering student at American International University-Bangladesh, and Co-Founder & Supervisor of Arché Intelligence Lab.
My research interests span Machine Learning, Deep Learning, Natural Language Processing, Computer Vision, and Bioinformatics, with a focus on solving real-world challenges in healthcare, online safety, and sustainable agriculture through intelligent systems.
Driven by curiosity and a passion for technology, I am dedicated to advancing innovation to address societal issues, from healthcare and environmental sustainability to online safety and bias mitigation.
Advanced expertise in ML algorithms (SVM, Random Forests, KNN) and NLP with transformer models (BERT, DistilBERT, XLM-RoBERTa). Hate speech detection, sentiment analysis, and text classification.
Human interaction object detection, medical image classification, and self-supervised learning approaches for visual understanding tasks.
Expert-level data manipulation with NumPy and Pandas. Creating impactful visualizations with Matplotlib, Seaborn, Plotly, and geographical plotting.
Building robust CI/CD pipelines for ML projects — from model training and validation to automated deployment and monitoring.
Full-stack development with C++, Python, JavaScript, PHP, and more. Building web applications, chatbots, and game development with OpenGL.
Applying computational methods to biological data, bridging the gap between AI research and healthcare innovation.
A friendly AI companion for image generation, text-to-speech, speech-to-text, and 3D image creation.
Self-supervised learning approach for classifying brain tumors from medical imaging data.
Hate speech and extremism detection using advanced BERT transformer models with high accuracy.
AI-powered chatbot that generates images using Stable Diffusion from text prompts.
2D game built with OpenGL and C++ featuring three levels, animations, physics, and audio integration.
Multi-variate signal processing and prediction system for trading bot applications.
End-to-end CI/CD pipeline setup for machine learning projects ensuring reliable deployment.
Object detection models developed for recognizing and analyzing human interactions in visual data.
Food delivery app built using JavaScript, PHP, HTML, and CSS. Features user management and database operations.
My research explores the intersection of machine intelligence and human language, focusing on solving real-world challenges in medical image classification, content moderation, sustainable agriculture, and bias mitigation. Several papers are currently under review in reputable journals.
T3SSLNet: Triple-Method Self-Supervised Learning for Enhanced Brain Tumor Classification in MRI - IEEE Access. DOI: 10.1109/ACCESS.2025.3589619
View PaperOpthaNet: Attention-Integrated Architecture for High-Precision Multi-Class Ophthalmic Image Classification - Healthacare Technology Letters, 2025, Wiley. DOI:https://doi.org/10.1049/htl2.70067
View PaperKbFL-XAI: Explainable Knowledge-Based Federated Learning for Eye Disease Diagnosis - Biomedical Engineering and Advances, Elsevier. DOI: 10.1016/j.bea.2025.100196
View PaperSSChNet: Self Supervised Learning based pre-trained network for Child Speech Recognition - Presented in NCIM 2025. DOI: 10.1109/NCIM65934.2025.11160221
View PaperVisionOcular: Transfer Learning with Self-Attention for Multi-Class Ocular Diagnosis - Presented in QPAIN 2025. DOI: 10.1109/QPAIN66474.2025.11171736
View PaperMIME: Multi-Strategy Imbalance Mitigation Engine for Skin Cancer Classification - Presented in TCCE'25, Melaka, Malaysia.
BUSegNet: A Speckle-Aware Multi-Scale CNN with Noise Modeling and Gated Calibration for Breast Ultrasound Diagnosis - Presented in ICCIT, Cox's Bazar.
Lightweight Supervised Approach for Behavioral Insider Anomaly Detection from SQL Logs - Presented in ICCIT, Cox's Bazar.
Parental Influence in Shaping Student Performance: A Data-Driven Study in Bangladesh - Presented in 7th IEOM International Conference, Dhaka, Bangladesh. DOI: BA07.20240171
View PaperCropGuardian: Revolutionizing Agriculture with AI-Driven Smart Farming Solutions - Presented in ICAMBE, BAU, Mymensingh, Bangladesh.
Ethic-BERT: An Enhanced Deep Learning Model for Ethical and Non-Ethical Content Classification - under peer review on AI and Ethics, SpringerNature.
Five consecutive academic sessions at AIUB, maintaining a high CGPA of 3.94.
Received 70% academic scholarship at AIUB based on merit.
Organized by Department of Physics, AIUB.
Biology Olympiad 2017: Organized by BDBO.
2017Organizer of National Ideas Challenge, Blood Donation, Winter Clothing Drive. HR at Oder Eid Amader Khushi(2025). Member of AIUB Robotic Crew & IEEE Student Branch.
Third Place Winner and Man of the Match in the university's flagship sports event (AWC'23).
2023I'm always open to discussing new projects, research collaborations, or opportunities to create meaningful impact through AI and technology.