Nozzle clogging is a major contributor to oxide inclusions in steel, impacting the cleanliness of the final product. These clogs frequently interrupt the casting process, resulting in downtime and diminished productivity. This project is dedicated to developing predictive models for anticipating nozzle clogging during continuous casting. By leveraging historical data and advanced analytics, operators can forecast potential clogging events, allowing for proactive interventions to minimize disruptions and optimize casting efficiency. Enhancing predictive capabilities in this area aims to improve steel cleanliness and overall productivity in continuous casting operations.