Yossra Noureldien CV - AI and Data Systems Lecturer PDF

Yossra Noureldien Experience AI and Data systems Lecturer Sep 2021 - Jul 2024 University of Khartoum, Sudan - Full time Redesigned...

Yossra Noureldien Experience AI and Data systems Lecturer Sep 2021 - Jul 2024 University of Khartoum, Sudan - Full time Redesigned and delivered 3 curricula for undergraduate courses in AI and expert systems, and database systems, resulting in around 40% increase in student satisfaction scores. Conducted 30+ interactive lectures, tutorials, and lab sessions, in- tegrating case studies and hands-on activities to enhance learning outcomes. Application Developer Mar 2021 – Jun 2021 Syber Technology, Sudan - Internship Summary Enhanced the Syber Delivery project by adding new features and managing API integrations. Experienced AI professional with 2 years of hands-on experience in artificial intelligence. Specializes in Education Natural Language Processing MSc in Telecommunications Engineering Jul 2020 (NLP), Generative AI, and Large University of Jaen, Spain Language Models (LLMs). Skilled in developing AI-driven solutions, BSc in Electrical and Electronics Engineering Nov 2017 with a strong focus on leveraging University of Khartoum, Sudan advanced models. Personal Projects Contact Details @ [email protected] LLM-Twin Project H +971 50 231 6401 Purpose: a personalized LLM Twin by using scraped data from personal m linkedin.com/yossra work, ensuring it reflects the user’s professional and personal expertise. B Itihad street, Sharjah, UAE Method: a RAG app with three pipelines workflow, the data pipeline scrapes and processes data; the model training pipeline fine-tunes Personal Information an LLM; and the inference pipeline for interaction and continuous monitoring. Citizenship: Sudan Age: 28 years Tools: ZenML for orchestrating ML pipelines, MongoDB for storing Languages: Arabic (native), scraped data, Qdrant as a feature store for vectorized representations, English (fluent), Spanish (B1) AWS SageMaker for LLM fine-tuning, Docker for containerization and Comet ML for experiment tracking and prompt monitoring. Skills Agent4Debate-AR: Autonomous Arabic Debate Agent LLM Techniques: RAG, Purpose: Extended an existing self-debate system inspired by the paper Agents, LLMOps "Agent4Debate",enabling support for Arabic language debates. LLM Frameworks: LangChain, Method: Adapted the original system by implementing an Arabic Comet ML, Zen ML, Hugging version of the debate flow. Face NLP Libraries: NLTK, SpaCy Tools: Utilized Autogen for agent development, Streamlit for interactive deployment, and LLMs for Arabic text generation. Applied custom ML Frameworks: Scikit-Learn, prompt engineering to align with Arabic linguistic nuances. TensorFlow Deployment & Cloud: AWS, Docker, Github Actions Web Tools: JavaScript, Streamlit, FastAPI Programming Languages: Python, SQL Version Control: Git