Jobayer is a CSE graduate from United International University (UIU), Bangladesh. He is a Research Engineer at the Advanced Intelligent Multidisciplinary Systems (AIMS) Lab at IRIIC at UIU, with strong experience applying AI to address critical societal challenges. He is adept at the full research lifecycle, from designing ML and LLM architectures and creating datasets to conducting literature reviews, data synthesis, and analysis. His work, including multiple peer-reviewed publications, has been accepted at top-tier venues such as IEEE ICDM and AAAI, and has contributed to securing $377,533 in research funding. His research interests include health informatics, human-centered computing, image processing, and computer vision. He has also served as an undergraduate teaching assistant at UIU.

Predictive Modeling for Primary Care Health Triage

The goal of this project is to reduce medication errors in low-resource clinical settings by developing an automated patient risk prediction and triage system. He curated a novel, clinically validated dataset of 130,637 patient scenarios, including structured patient data such as age, gender, symptom duration, severity, and 1,816 unique symptoms, where each case was reviewed by a panel of expert physicians to ensure clinical accuracy. Using PyTorch and BERT-based architectures, he designed a benchmark and optimization framework comparing ensemble ML and fine-tuned LLM approaches for automated health triage. The framework was benchmarked for risk-stratification accuracy and precision, and was accepted for publication at IEEE ICDM 2025 in Washington, DC, USA, and the 40th AAAI 2026 Conference in Singapore. In future work, he aims to integrate real-time physiological data, address demographic bias, and is currently conducting a pilot study to test system adaptivity.

SmartAMR: Advancing Digital Healthcare with AI-driven Intelligence 

As a key contributor to this World Bank-funded project, he is working on developing an AI-based intelligent system to reduce antimicrobial resistance (AMR) in Bangladesh. The project integrates an AI-driven symptom checker, real-time prescription auditing across drug classes, and medication adherence monitoring into a cohesive healthcare ecosystem. The system supports hospitals, pharmacies, and community health workers by ensuring evidence-based prescribing and reducing prescription errors through AI decision support integrated into EMR platforms.

Explainable AI (XAI) for Diabetes Prediction

To address clinical skepticism toward black-box models, he developed a transparent prediction framework for Type-2 Diabetes. He created the "DiaBD" dataset comprising 5,437 patient records in collaboration with community health workers and implemented a feature selection strategy using Scikit-learn and XAI libraries, including SHAP and LIME, to provide transparent and trustworthy risk assessments for healthcare providers. The dataset was published in Data in Brief, and the model paper is currently under review. Future improvements include expanding data collection to reduce demographic imbalance.

1. T.T.Prama, M.J.Rahman, M.Zaman, F.Sarker, & K.A.Mamun, (2025). DiaBD: A Diabetes Dataset for Enhanced Risk Analysis and Research in Bangladesh. Data in Brief, 111746.
   DOI: 10.1016/j.dib.2025.111746
2.  M.J. Rahman, S. Ahamed, F. F. Deepto, M. Zaman, A. Islam, F. Sarker, & K. A. Mamun, (2025). A Benchmark and Optimization Framework for ML and LLM-based Health Triage. Accepted to BPOD @ ICDM 2025 (IEEE International Conference on Data Mining 2025 Washington DC, USA).
3. E. Ghosh, M. J. Rahman, S. Ahamed, M. Salwa, T. Foyez, K. A. Mamun, (2026). Integrating AI-Driven Triage into Digital Pharmacy Systems for Rational Antibiotic Use in Low-Resource Settings. AAAI Bridge @ AAAI-26 (40th AAAI Conference on Artificial Intelligence- Singapore).