When AI Can Do the Analytics, What Remains to Be Engineered?

Speaker 

Dr. Li-C Wang is a professor at the ECE department at the University of California, Santa Barbara.

Abstract

Large Language Models are rapidly changing how semiconductor data analytics can be performed. Modern AI systems can generate code, analyze tables, create visualizations, and autonomously carry out multi-step analytic tasks. This raises a fundamental engineering question: if AI can increasingly perform the analytics, what remains to be engineered?

This talk argues that the challenge is shifting from implementing individual analytic algorithms toward engineering the environment in which AI agents operate. Semiconductor data carries domain semantics, organizational conventions, workflow assumptions, and engineering knowledge that are not inherently understood by a general-purpose AI model. Reliable agentic analytics therefore requires mechanisms for grounding data, capturing engineering knowledge, and providing appropriate context for AI reasoning.

I will present IEA-Connect, an architecture developed around this perspective. It combines workflow-driven knowledge acquisition with autonomous agentic analytics through Analytic-Ready Tables (ARTs), which provide grounded reasoning environments for AI agents. A key idea is that validated engineering solutions themselves become reusable context: one engineer’s solution can become another engineer’s—or another AI agent’s—starting point. The talk will discuss the resulting primal–dual paradigm, the changing role of engineers, and practical design principles for building scalable and reliable AI-native analytics systems for semiconductor engineering.

Biography

Li-C. Wang is a Professor in the Electrical and Computer Engineering Department at the University of California, Santa Barbara. He received his Ph.D. from the University of Texas at Austin and previously worked at Motorola’s PowerPC Design Center. Since 2003, his research has focused on applying machine learning and AI to semiconductor design and test flows, resulting in over 100 papers and 22 supervised Ph.D. theses.

He has received 12 Best Paper Awards—including recent awards at the International Test Conference (ITC) in 2024–2022 and 2020—and the 2010 SRC Technical Excellence Award for his work in data mining for test and validation. He is also the recipient of the IEEE TTTC Bob Madge Innovation Award (2017). An IEEE Fellow, he served as Program Chair (2016) and General Chair (2017, 2018, 2023) of ITC.