Cascaded Microring Resonators (CMRR), a new type of optical sensor, are widely used in biology, medicine, and other fields because of their high sensitivity, easy integration, and low power consumption. In this paper, we propose a Python-based envelope fitting method for real-time CMRR sensor’s output spectrum to achieve real-time data analysis and processing of the CMRR sensor’s output spectrum. First, different fitting models were used to fit the output spectrum of the CMRR sensor. Then, the fitting errors of different fitting models were compared by sensitivity error percentage, and it was concluded that the smooth spline fitting method performed best in real-time processing of the output spectrum of the CMRR sensor. Finally, NaCl solution with different concentrations was used for real-time acquisition and processing of the output spectrum. The reliability of the real-time acquisition and processing program for the CMRR sensor’s output spectrum is verified. The experimental results show that the wavelength drift of the CMRR sensor is linearly related to the concentration of the solution. It can be seen from the calculation that the sensitivity of the CMRR sensor for brine is about 671.03529 nm/RIU.