Ahmed Abdullah
ahmedembedded [@] gmail.com
l227503 [@] lhr.nu.edu.pk

I am an undergraduate researcher in machine learning and artificial intelligence at the National University of Computer and Emerging Sciences (FAST-NUCES), where I work on machine learning across multimodal learning, medical imaging, and natural language processing.

My research focuses on building robust and explainable AI systems, with applications in radiology, multilingual NLP, and cross-modal representation learning. I have collaborated with clinical and academic institutions on projects in medical imaging and multimodal AI, and have contributed to peer-reviewed publications in these areas.

Alongside my academic work, I have worked on production-scale machine learning systems in healthcare and conversational AI, bridging research and real-world deployment.

CV  /  Resume  /  Google Scholar  /  LinkedIn  /  GitHub  /  Undergraduate Thesis

My research focuses on multimodal machine learning, medical imaging, and natural language processing. I am particularly interested in building robust and explainable AI systems that operate under limited data and cross-modal settings, including face-voice association, radiological imaging, and multilingual language understanding. Recently, I have also been working on scalable data systems and real-world deployment of machine learning models in healthcare settings. For a full list of publications please see Google Scholar.

Large-scale evaluation and analysis of multilingual face-voice association systems under cross-lingual conditions.

Benchmark design and evaluation framework for multilingual multimodal identity association.

A multilingual transformer-based framework for hope speech detection across low-resource languages.

An empirical analysis of publication trends, retractions, and research integrity patterns in Pakistan.

Pseudo-multimodal learning framework for lumbar spine degeneration classification using radiological imaging.

GAN-based augmentation framework for thoracic disease recognition under limited data regimes.

An explainable multilingual phishing detection system using transformer-based language models.

This guy is good at website design.