Modern manufacturing systems are equipped with a multitude of sensors that continuously generate extensive time series data capturing the condition of production processes. A constant challenge is linking time series data to spatial product features or specific process states. The automated tape laying process exemplifies this challenge: defects are detected with high spatial resolution on the product surface, yet the defect origins must be traced back to corresponding segments in the time series data. Such a link can be achieved using process models. Process models describe the causal relationships of an underlying production process by connecting sensor signals with operating principles and provide these in the form of refined data. Beyond enabling the assignment of time series data to spatially resolved component features, process models also facilitate process analysis and optimization. Especially in data-driven modeling approaches, the core algorithms are often identical, while the resulting structure, parameterization, input, and output data vary depending on the application. Managing these models and their parameters in a systematic and reusable way is therefore critical for efficient process analysis and optimization. This paper proposes an extension of an existing process database to store process models along with their underlying algorithms. The approach aims at a semantic link between input and output data, the specific parameterization of a model, and the corresponding process context. By adhering to the FAIR principles, the approach not only enables effective model reuse and reproducibility but also facilitates the contextualization of parameterized models and their results within the underlying production processes.