Humans have created over 200 million different chemicals, but we only have comprehensive property data for less than 0.05 % of them. This gap in data availability hinders progress in the chemical industry, such as developing novel products and supply chains essential for the transition to a sustainable carbon economy. The demand for physical property data required for the transition can only be served in time with accurate property prediction methods. Our team, Philipp, Gernot, Johannes, Benedikt, and André, addresses this problem with an advanced machine-learning-powered property prediction tool that integrates natural language processing with modern physical property models. Our user-friendly and comprehensive solution is designed to meet the needs of customers in the chemical industry.