About the Speaker

Abdulrahman Elsadiq is a Machine Learning Engineer at Alpinum Consulting, specializing in the application of AI to hardware design and verification. His work focuses on integrating generative AI, machine learning, and agent-based systems into RTL and DV workflows to improve productivity and scalability. He has experience in ML pipelines, LLM fine-tuning, and digital design verification, including FPGA-based AI acceleration and automated verification environments. Abdulrahman is passionate about bridging advanced AI techniques with practical semiconductor engineering challenges.

Agentic AI in Design Verification: From Assistance to Autonomous Workflows

Overview

Agentic AI is emerging as a transformative paradigm in design verification, introducing systems capable of autonomously orchestrating complex verification tasks. This presentation explores how AI-driven agents can be integrated into existing DV environments to enhance testbench generation, accelerate debugging, and improve coverage efficiency. Rather than focusing on theory, the session highlights a concrete workflow and its practical impact on engineering productivity. Key challenges, including reliability, control, and integration, are discussed, along with strategies for adopting AI in a controlled, human-in-the-loop manner within production verification flows.

Key Points

• Agentic AI in DV: How intelligent agents differ from traditional automation and where they fit in modern verification workflows
• Practical Workflow: Applying AI to testbench generation, debug, and coverage improvement within existing DV environments
• Adoption Strategy: Key challenges, limitations, and how to safely integrate AI using human-in-the-loop approaches