Zahrah Rasheed
Artificial Intelligence (Fresh Graduate)
Artificial Intelligence graduate passionate about building intelligent, data-driven solutions and applying machine learning to solve real-world problems.
About Me
Hi, I’m Zahrah Rasheed, a fresh graduate with a Bachelor’s degree in Artificial Intelligence (GPA 4.53/5.0 – Second Class Honors) from the University of Jeddah. I am passionate about building intelligent systems that solve real-world problems.
I specialize in machine learning, data preprocessing, and model optimization, with hands-on experience using Python, PyTorch, Pandas, and Power BI. My work includes developing deepfake voice detection systems, applying NLP in healthcare contexts, and improving data pipelines to enhance AI model performance. I enjoy transforming complex data into meaningful, actionable insights.
My projects combine analytical thinking with technical precision, including classification models, Django-based web applications, and interactive Power BI dashboards. I am particularly interested in deep learning, language models, and intelligent automation.
Currently, I am expanding my expertise by working on AI-enhanced systems across various domains, aiming to deliver innovative and impactful AI solutions.
Solo Projects
Professional Certificate – NVIDIA Certified Associate: Generative AI (LLMs)
Earned NVIDIA’s Generative AI certification, validating knowledge in Large Language Models (LLMs), transformer architectures, prompt engineering, and AI deployment concepts. This certification strengthens my expertise in modern AI systems and generative technologies.
View Credential: https://www.credly.com/earner/...
A Deep Learning Framework for the Detection of DeepFake Audio IEEE 3rd International Conference on Business Analytics for Technology and Security (ICBATS), 2025
Publication
Co-authored and published research introducing deep learning models for detecting deepfake audio. The study evaluated CNN, BiLSTM, and hybrid architectures using spectrogram-based feature extraction and data augmentation techniques. The optimized CNN model achieved 93.54% accuracy in distinguishing real and synthetic audio.
The project also included the development of a web-based prototype for real-world application of the detection system.
🔗 View Publication on IEEE Xplore: https://ieeexplore.ieee.org/do...
Achievements
Award for Digital Innovation – University of Jeddah & HOTS Academy
Awarded for the Tafahas Project, at the University of Jeddah Digital Business Incubator Event. Selected as one of 11 projects recognized for innovation and impact in digital security and AI entrepreneurship.
