For such a process calibration installation, an online GC calibrated to the analytes in the stream as well as the sampling manifold is integrated to the process downstream of the Raman Analyzer to ensure that both the online GC as well as the Raman analyzer measure the same sample. When a measurement cycle is completed, the Raman spectra and corresponding primary analysis data from the GC are pushed to an AI enhanced cloud modelling environment. The model is initially created by the chemometric platform, and all subsequent measurement cycles by GC/Raman are then used to enhance the model. The model is then loaded into the Raman analyzer software, using these data points to improve the subsequent predictive performance of the Raman system. After as little time as a few weeks, the online GC can be removed from the calibration process. The Raman analyzer can then be connected to the control system, DCS/PLC, to control relevant process controls e.g. steam valves, flow rates, feedstock. After about 3 months, this process will reveal savings from real-time process monitoring and control. At this point, the offline lab measurements are no longer necessary to monitor the process.