Muhammad Abdullah

Led multiple AI/ML research projects with a combined funding portfolio of €1M+, driving technical strategy, project execution, and team coordination. Supervised interdisciplinary AI teams and directed computer vision and hyperspectral image analysis in collaboration with the University of Birmingham's BIFoR-FACE project.

Developed an AI-powered bilingual (Urdu/English) vocational learning platform for Auto-Electrician programs, supporting personalized learning, content generation, and learner assessment. Engineered an LLM-based Q&A pipeline using LLaMA-13B/70B, RAG, and vector databases, generating 48,000+ Q&A pairs.

Developed an adaptive multilingual learning platform using LLM-based personalization for vocational education. Conducted R&D on LLM-driven adaptive learning systems, including RAG pipelines, multi-agent architectures, and personalized content generation.

Built a machine learning-based browser security system that classifies URLs in real-time and prevents access to malicious websites, achieving 0.97 F1-score and 0.94 precision on the minority class. Addressed severe class imbalance and built an end-to-end threat detection system using Chrome Extension, Flask REST API, and MongoDB.

Ongoing certification in Introduction to Back-End Development from Meta.

Ongoing certification in Introduction to Front-End Development from Meta.

Ongoing certification in Supervised Machine Learning: Regression and Classification from DeepLearning.AI.

Activities and societies: GPA Scholarship, Publication Scholarship, DAAD Scholarship

Bachelor of Science in Computer Science from National University of Sciences & Technology, with scholarships for academic excellence and research publications.