Yahya Iranmanesh

Instructor at Azad University and Ph.D. candidate in Computer Engineering (Computer Systems Architecture).

He has conducted advanced research and designed novel hardware accelerators for neural networks, focusing on optimizing computational performance, minimizing power consumption, and enhancing scalability for complex AI models. His work enables the efficient deployment of AI-embedded systems in a variety of real-time applications, bridging the gap between high-performance neural network algorithms and resource-constrained hardware platforms. He has developed FPGA-based prototypes, designed custom Neural Processing Units (NPUs), and implemented AI models directly on embedded devices, enabling applications such as intelligent robotics, machine vision systems, industrial automation, and autonomous control systems. Furthermore, he explores novel architectures for hardware-aware neural network optimization, including parallelization techniques, quantization strategies, and low-latency inference pipelines to maximize throughput while maintaining accuracy. His research also integrates AI model design with hardware-software co-design methodologies, ensuring that the final embedded system achieves both efficiency and reliability in practical industrial and research environments.

In the industrial sector, he works on designing fully automated production lines and converting manual machines into automated systems for various companies.

Additionally, he develops and implements dedicated software for full office automation platforms across multiple organizations. These platforms are AI-driven and incorporate intelligent process automation, digital assistant systems, and tools for enhancing customer relationship management, sales optimization, and workflow efficiency. By integrating machine learning and AI algorithms, these systems enable organizations to automate repetitive tasks, provide intelligent recommendations, and improve overall operational productivity and decision-making processes.