Dubai, United Arab Emirates
I am a recent Computer Science graduate (GPA 3.84) specializing in Artificial Intelligence, Machine Learning, and Natural Language Processing. Throughout my academic and research journey, I have been deeply focused on building AI systems from the ground up, combining theoretical rigor with practical engineering. My work spans across model development, optimization, and large-scale deployment. At the Technology Innovation Institute (TII), I contribute to the advancement of the next generation of Falcon models, focusing on multi-modal (vision and text) capabilities and enhancing reasoning abilities. Beyond my formal work, I have independently recreated architectures like LLaMA 3.2 and GPT-2 from scratch in PyTorch, building my own custom tokenization pipelines and successfully exceeding baseline performances. I have also recreated post-training reasoning optimization frameworks such as GRPO (Group Relative Policy Optimization) to strengthen LLM reasoning capabilities after pretraining. I have also developed multi-node training setups to scale LLM training efficiently. Some other achievements include: • Fine-tuning LLaMA-based models for software vulnerability detection using LoRA, DoRA, and federated learning techniques. • Leading an award-winning research project deploying Retrieval-Augmented Generation (RAG) with long-context support, recognized by Huawei and TII. • Engineering forecasting pipelines that outperformed traditional ML models in the energy sector during my internship at AIQ (a G42 company). • Researching prompt engineering strategies to maximize zero-shot and few-shot LLM capabilities across applied projects. • Building security-focused AI models, including URL risk classifiers and vulnerability repair tools. I am passionate about scalable, intelligent AI systems that bridge theory and real-world application. Currently, I am seeking opportunities to contribute to innovative AI research and development teams pushing the boundaries of what LLMs and multi-modal AI can achieve. Key Skills: • AI/ML model development and deployment • LLM fine-tuning (LoRA, DoRA, federated learning) • Retrieval-Augmented Generation (RAG) systems • Multi-modal AI (vision + text integration) • Distributed training and multi-node system setups • Prompt engineering and zero-shot optimization • Programming: Python, PyTorch, TensorFlow
Working on LLMs and Data.