Muhammad Talha Hassan

PROFESSIONAL PROFILE I am a Final-year BS Electrical Engineering student at NUST with a minor in Artificial Intelligence and a strong interest in machine learning and deep learning research. My academic and lab experience includes working on large language models, reinforcement learning, and multimodal signal processing. I enjoy building practical ML systems that combine solid theory with real-world impact. Currently, I am focused on expanding my research skills and contributing to data-driven and AI-focused projects. EDUCATION BS Electrical Engineering SEECS , Islamabad , 3.66 (2026) INTERNSHIP EXPERIENCE Deep Learning Lab SINES, NUST 01-Jun-2025 - 31-Aug-2025 Gained hands-on experience with LLaMA- and DeepSeek-style transformer architectures for large language models, including the implementation and analysis of advanced components such as Grouped Multi-Query Attention, Rotary Positional Embeddings (RoPE), and SwiGLU activations. Worked with state-of-the-art models including LLaMA-2-7B, LLaMA-4-17B-16E, DeepSeek-V3, and DeepSeek-R1 to study modern reasoning-oriented LLM designs. Implemented the LLMCompass framework for FPGA deployment on the Xilinx ALVEO platform, providing simulation, performance analysis, and documentation to support efficient hardware acceleration of large language models. MachVis Lab SEECS, NUST 01-Jun-2023 - 31-Aug-2023 This internship focused on geospatial data processing and machine learning fundamentals. Developed a Python-based pipeline to convert point-based geospatial data into polygonal boundaries using Shapely, implementing convex hull and alpha shape algorithms for accurate region delineation. Built interactive geographic visualizations using Folium to effectively represent spatial insights. FINAL YEAR PROJECT Ghaat: AI-Powered Autonomous Aerial Security This project focuses on the design and development of an autonomous hexacopter interceptor capable of stable flight with real-time detection, tracking, and interception of rogue drones. It aims to implement onboard AI and edge computing for processing live video feeds to enable target identification and trajectory planning. The system integrates stereo vision to achieve accurate depth perception and localization. Research conducted on an autonomous capture mechanism to neutralize or secure target drones without human intervention. TECHNICAL EXPERTISE Artificial Intelligence/Machine Learning Strong foundation in machine learning and deep learning with hands-on experience in Transformer-based architectures, Large Language Models, and reinforcement learning. Worked on computer vision and multimodal learning problems involving attention mechanisms, representation learning, and real-world datasets, w ...

I am a Final-year BS Electrical Engineering student at NUST with a minor in Artificial Intelligence and a strong interest in machine learning and deep learning research. My academic and lab experience includes working on large language models, reinforcement learning, and multimodal signal processing. I enjoy building practical ML systems that combine solid theory with real-world impact. Currently, I am focused on expanding my research skills and contributing to data-driven and AI-focused projects.