At Outlier, I tested and evaluated pre-trained AI models for performance, accuracy, and stability, then reported bugs and edge cases with structured feedback to improve model reliability. As an IBM Data scientist Intern at DEPI, I prepared and analyzed datasets, applied machine learning algorithms, and created visualizations and dashboards for data-driven projects.
For my graduation project, I fine-tuned OpenAI's Whisper for Arabic Quran recitation transcription, improving accuracy for clean speech data. I also built a RAG-powered PDF question answering system and developed machine learning and computer vision projects, including a YOLOv11 system for detecting safety equipment.

