This study presents the design, implementation, and validation of GluFi, a non-invasive glucose meter based on Near-Infrared (NIR) reflectance spectroscopy, with a primary focus on analog and digital signal processing and sensor system design. The analog front-end was developed using active Sallen-Key filters and a first-order low-pass filter to minimize electromagnetic interference and external noise, ensuring stable and accurate signal acquisition. Two infrared LEDs operating within the 940–960 nm range were precisely aligned with a sensitive phototransistor to target glucose-specific absorption in the capillary bed (while measuring on finger) or radial artery (when measuring on the wrist). Optical pathway calculus guided the sensor placement to maximize measurement sensitivity while reducing interference from superficial tissues. Signal acquisition was performed with an ESP32 microcontroller paired with an ADS1115 analog-to-digital converter, offering high-resolution digitization. After digitization, raw sensor data underwent soft digital filtering and baseline correction to improve signal-to-noise ratio. A calibration algorithm was implemented to correlate processed signals with reference glucose values obtained from a standard invasive glucometer. Experimental validation against a commercial glucometer demonstrated promising clinical accuracy, though identified areas for further refinement in filtering parameters and calibration procedures. Future work will concentrate on optimizing both analog and digital processing stages, refining sensor geometry, and enhancing calibration methods to advance GluFi toward clinical applicability.