AI enabled process control is helping foundries and automakers eliminate defects in completed products through the use of dynamic machine learning models. Predictive and prescriptive AI provide the necessary limits to control variables to preemptively avoid defects at every point of the production process, enabling real-time root cause analysis and corrective action suggestions. These corrective actions help foundries and automakers avoid cascading defects that previously resulted in costly scrapping, rework, and downtime. The implementation of AI process control translates into significant savings and ROI for heavy manufacturing companies. We will relate our experience in implementing AI for process control in the casting and automotive assembly sectors – both the successes and challenges we have faced.