Limited access to analytical instrumentation motivates the development of affordable alternatives. This study describes the construction and evaluation of two ESP32-based optical prototypes: an RGB colorimeter using a TCS34725 sensor and an ultraviolet radiometer using an AS7331 sensor. The RGB device combined a darkened chamber with a mirror-assisted optical path and was evaluated using colored papers and solutions of potassium permanganate, methylene blue, rhodamine B, zinc phthalocyanine, and free-base tetraphenylporphyrin. Paper measurements distinguished colors but differed from nominal RGB codes due to factors as print configuration and paper type. Comparisons with a benchtop spectrophotometer revealed analyte- and channel-dependent linear ranges, with R² values of 0.950-0.999 for selected RGB fits. In a test using an unfocused UV flashlight, the recorded signal decreased with increasing source-sensor distance, exhibiting behavior close to that predicted by the inverse-square law. Solar monitoring yielded temporal profiles qualitatively similar to commercial records obtained at a different location and date. Dedicated applications supported data acquisition, visualization, and storage. Artificial intelligence assisted development under human supervision. The results indicate potential for accessible instrumentation, experimental teaching, and connected monitoring, while emphasizing application-specific calibration and the limitations of the evaluated configurations.