Built code-focused conversational datasets for LLM training, specializing in Python, data analysis, and Machine Learning.Crafted realistic technical dialogues reflecting how programmers discuss code, solutions, and best practices.Ensured data accuracy and relevance, contributing to high-quality LLM training for code understanding and generation.Utilized Python and Jupyter notebooks to structure, document, and optimize code-based conversational data.Evaluating the quality of AI-generated code, including human-readable summaries of your rationale.

Developed an Arabic question-answering system using AraBERT, a state-of-the-art NLP model specifically designed for the Arabic language.Implemented a multi-step process including question processing, context retrieval, information search, and answer extraction to accurately respond to user queries.Utilized advanced NLP techniques such as tokenization, named entity recognition, and attention mechanisms to enhance the model's understanding and performance.Leveraged AraBERT's pre-trained knowledge on 3B words (70M sentences) to achieve superior context understanding and linguistic relationship comprehension in Arabic.Compared model performance against alternatives like AraGPT, demonstrating AraBERT's advantages in bidirectional context understanding for question-answering tasks.

Analyzed Facebook ad campaign data for GlobalShala's Superhero U event, creating impactful visualizations of key performance metrics.Developed skills in data visualization techniques, applying principles of visual thinking to effectively communicate complex insights.Contributed to team efforts in evaluating campaign performance and making data-driven recommendations for resource allocation.Designed audience-appropriate data presentations, adhering to best practices in data visualization for marketing analytics.Collaborated in a diverse, remote team environment, enhancing data analysis, critical thinking, and virtual communication skills.