Michael B. Khani (b. 1995) is a seasoned computer engineer and artificial intelligence researcher with over six years of experience in software engineering and automated systems. Currently pursuing his Ph.D. in Artificial Intelligence at the University of Göttingen in Germany, he is recognized for his groundbreaking research in scalable AI systems, cyber-physical systems, and information theory. Michael holds a Master’s degree in Computer Engineering from Karabuk University, Turkiye, and a Bachelor’s in Computer Software Engineering from Rasht Ahrar University in Iran.
A distinguished scholar, Michael has been twice honored with the prestigious Young Scientist Award by the Presidential Scientific and Research Deputy of Iran, in 2017 and 2023. He is an active member of the ACM and its Special Interest Group on Artificial Intelligence (SIGAI), and also belongs to Iran’s National Elites Foundation. Professionally, he serves as a Research Assistant and Lecturer at the Institute of Computer Science at the University of Göttingen, and consults independently as a machine intelligence specialist.
His extensive publication record includes peer-reviewed contributions in IEEE Access, COMPSAC, and IARIA conferences, covering topics such as neural attention models, AI-driven predictive maintenance, and scalable neural network architectures. He has also co-submitted patents related to neuro-prosthesis design and self-regulating electronic circuits.
Michael's skill set spans artificial intelligence, machine learning, software engineering, and cognitive robotics. Fluent in English and Persian, with intermediate proficiency in Turkish and basic German, he is known for his adaptability, strong problem-solving abilities, and collaborative spirit. His ultimate goal is to pioneer explainable AI solutions for high-performance computing environments.
AI systems, computational intelligence, and cognitive robotics. He also focuses on cyber-physical systems, emergent intelligence, and information theory.
Towards Scalable AI Systems: Predictive Maintenance and Federated Intelligence Across the Compute Continuum
As digital infrastructure evolves into a heterogeneous ecosystem of edge, cloud, and high- performance computing (HPC) systems, ensuring reliability and resilience is more critical than ever. This talk introduces the novel today's advances that unify predictive maintenance with federated learning to enable scalable, privacy-preserving intelligence across distributed compute environments. Drawing from real-world HPC logs, AI workload quantization models, and LLM-enhanced anomaly detection, I will present how modern systems can self- monitor, self-learn, and adapt to changing workloads with minimal downtime. The presentation also highlights ongoing efforts to bridge research with practical deployment in critical infrastructure domains.