Kinetic model discovery in chemical engineering has been hindered by the lack of accurate rate expressions, but a new approach combines domain-aware symbolic regression with large language models (LLMs) to identify interpretable kinetic models. This method, known as DASyR-LLM, leverages domain knowledge to guide the search for plausible models, addressing a key limitation of traditional symbolic regression techniques. By incorporating physicochemical constraints, DASyR-LLM can explore a more focused subset of possible models, increasing the likelihood of discovering accurate and meaningful rate expressions1. The integration of LLMs enables the approach to learn from large datasets and generate models that are consistent with domain expertise. This development has significant implications for the field of chemical engineering, as it can lead to improved process control and optimization. The ability to accurately model kinetic processes can also have far-reaching consequences for related fields, such as materials science and biotechnology, so practitioners should pay close attention to the potential applications and limitations of DASyR-LLM.