Björn Schuller, Adria Mallol-Ragolta, Alejandro Peña Almansa, Iosif Tsangko, Mostafa M. Amin, Anastasia Semertzidou, Lukas Christ, and Shahin Amiriparian. “Affective Computing Has Changed: The Foundation Model Disruption” In: arXiv preprint arXiv:2409.08907, 2024.
Mina A. Nessiem, Harry Coppock, Mostafa M. Mohamed, and Björn W. Schuller. "Artificial Intelligence in COVID-19". In: "Omics Approaches and Technologies in COVID-19", 2022.
Hannah Bast, Matthias Hertel and Mostafa M. Mohamed. "Tokenization Repair in the Presence of Spelling Errors“. In: Proceedings of the 25th Conference on Computational Natural Language Learning. Online: Association for Computational Linguistics, 2021, S. 279–289.
Alexander Nutz, Daniel Dietsch, Mostafa Mahmoud Mohamed and Andreas Podelski. "Ultimate kojak with memory safety checks". In: International Conference on Tools and Algorithms for the Construction and Analysis of Systems. Springer. 2015, S. 458–460.
Education
Ph.D. degree, 2024, Chair of Embedded Intelligence for Health Care and Wellbeing, University of Augsburg, Augsburg, Germany. Supervisor Prof. Dr. Björn W. Schuller
Dissertation title: "Foundation Models in Affective Computing". Submitted, pending defence.
Conducted peer-reviewed studies for using foundations models and LLMs (OpenAI GPT-4 and GPT-3.5) for more than 13 affective computing problems, including sentiment analysis and emotions detection and prompt engineering.
External researcher at the Chair of Health Informatics, Technical University of Munich (TUM), Prof. Dr. Björn W. Schuller, 2024.
M.Sc. degree in Computer Science, 2018, Alberts-Ludwig Universität Freiburg, Freiburg, Germany.
Chair of Algorithms and Data Structures, Prof. Dr. Hannah Bast
B.Sc. degree in Computer Science and Engineering, 2014, German University in Cairo (GUC), Cairo, Egypt.
Work Experience
Dynatrace: Senior Researcher in Generative AI, Nov 2024 -- Now, Vienna, Austria
SYNCPILOT: Senior Research Data Scientist, Dec 2017 -- Oct 2024, Germany
Led the development of LLM-based solutions for developers, creating generative AI interfaces (Django web platform, VSCode extension), implementing advanced RAG techniques for deep code understanding (syntax-based splitting, vector indexing), and building LLM agents for workflow automation.
Led the development of a real-time social signals prediction product, utilizing deep learning (Keras/TensorFlow) for multimodal personality prediction (audio/images/text), designing streaming solutions for live predictions (Redis, NVIDIA Triton), deploying scalable MLOps infrastructures (Kubernetes, Docker, Kubeflow), and mentoring a team on ASR and sentiment analysis.
University of Freiburg: HiWi Job, Summer 2016 - Autumn 2017, Freiburg, Germany
Freiburg RNA Tools: a tool that visualizes several RNA interaction algorithms for teaching purposes.
Google: Software Engineering Intern, Winter 2016, Paris, France
Facebook: Software Engineering Intern, Fall 2015, New York City, US
German University in Cairo: Junior Teaching Assistant, Spring 2014, Cairo, Egypt
Teaching computational number theory in the course " Advanced Data Structures and Algorithms ", slides: Lecture 1 and Lecture 2
Trustious: Research and Development Intern, Summer 2012, Cairo, Egypt
Competitive Programming
As Judge:
Chief Judge, Africa and Arab Collegiate Programming Contest (ACPC) 2018 and 2019, Sharm El-Sheikh, Egypt