PhD Candidate at Missouri S&T · Research Intern at Argonne National Laboratory
I build robust and trustworthy ML/LLM systems, with a focus on black-box/zero-access adversarial attacks, LLM security, and unsupervised representation learning.
Hello, I’m Nuzaer Omar, a doctoral student in Computer Science at Missouri University of Science and Technology (Missouri S & T), where I conduct research in the Wireless to Cloud Computing (W2C) Lab under the guidance of Dr. Sanjay Madria. I am currently completing the coursework phase of my PhD program, with an expected graduation date of July 2027.
My research is driven by the goal of building robust, reliable, and trustworthy NLP systems, with a particular focus on large language models (LLMs). I am especially interested in understanding and improving LLM behavior under adversarial and failure prone conditions, including black-box and zero-access attack settings that closely resemble real world deployment scenarios. My work emphasizes not only model accuracy, but also robustness, interpretability, and reproducibility, which I believe are essential for deploying LLM-based systems responsibly.
My interest in machine learning and AI emerged during my undergraduate studies in Electrical and Electronic Engineering at Chittagong University of Engineering & Technology (CUET). For my undergraduate thesis, I worked on the automated detection of cardiovascular diseases from ECG signals, where I combined deep learning and signal processing techniques to design interpretable diagnostic models. This work resulted in highly accurate yet low complexity models and later evolved into multi-class classification systems capable of distinguishing heart attack patients, other cardiac conditions, and healthy individuals. Through this process, I gained early exposure to data centric modeling, feature engineering, and the importance of domain aware interpretation principles that continue to shape my research today.
After completing my undergraduate degree, I joined Port City International University (PCIU), Bangladesh, as a Lecturer in the Department of Electrical & Electronic Engineering. In addition to teaching courses such as programming, electronics, signals and systems, and control systems, I supervised undergraduate student teams and mentored them in idea competitions and applied projects. These teaching and mentoring experiences strengthened my ability to communicate complex technical concepts clearly and reinforced my commitment to research that addresses real-world, human centered problems.
At Missouri S & T, I have continued to broaden my academic and professional engagement beyond core research. I have mentored undergraduate researchers, participated in research translation and commercialization initiatives such as NSF I-CORPS, and contributed to the scholarly community by serving as a peer reviewer for conferences including IEEE BigData and ECAI. These experiences have helped me develop a balanced perspective that bridges theoretical research, system-level implementation, and practical impact.
Looking ahead, my long term objective is to pursue a leadership role in applied research on large language systems, either in academia or in an industrial research environment. I aim to design AI systems that are not only powerful, but also safe, resilient, and aligned with human decision making, particularly in high stakes or uncertain settings. Ultimately, I hope my work will contribute to reducing cognitive burden for practitioners and improving the reliability of AI assisted decision support systems.
Conducting research on machine learning and signal-processing methods for gamma-ray source detection, anomaly identification, and robust multimodal modeling.
Conducting research on robust and reliable language models, adversarial attacks and defenses, unsupervised text clustering, representation learning, and computer vision.
Taught programming, electronics, control systems, microprocessors, measurement and instrumentation, and related laboratory courses. Supervised student projects and mentored teams in national competitions.
LLM Security · Adversarial NLP · Black-Box Evaluation
Unsupervised Learning · Representation Learning · Robust NLP
Unsupervised Clustering · LLM Reasoning · Semantic Representation
Computer Vision · Adversarial ML · Thermal Imaging
Argonne National Laboratory · Signal Processing · Transformer Models
Biomedical Signal Processing · ECG · Deep Learning
NLP · LLM Security · Robustness Evaluation
Unsupervised Learning · Clustering · Representation Learning
Detection · Segmentation
These are course-materials from my prior lecturer role at Port City International University (PCIU).
Python, MATLAB, C/C++, R, Verilog, x86 Assembly
PyTorch, TensorFlow, Hugging Face, scikit-learn, OpenAttack, Pandas, OpenCV, spaCy, NLTK, Ollama, vLLM
SQL, MySQL, MongoDB, Redis
LaTeX, Overleaf, Git, Docker, Slurm, GNS3, Proteus, Cadence Virtuoso, PSpice
CNNs, Autoencoders / VAEs, Contrastive Learning, Representation Learning, YOLOv5 / YOLOv8, Clustering
Transformer Models (BERT, T5, GPT, BART, LLaMA), Adversarial Attacks & Defenses, Prompt Engineering, Decoding Strategies, Semantic Similarity (SBERT), Text Generation, Topic Modeling, Phonetic & Lexical Perturbations
Loss Design (Contrastive, Reconstruction, Triplet, KL), Hyperparameter Tuning, Model Calibration
Email: nom8m@umsystem.edu · nuzaeromar97@gmail.com