This paper presents CABRA (Computer Automated Brainstem Response Audiometry), a low-cost, modular system designed to perform auditories through auditory brainstem response analysis. It relies on an ESP32-S2 microcontroller to acquire and digitize physiological signals from electrodes placed on the scalp. These signals, filtered and amplified, are processed by desktop software that generates auditory stimuli and analyzes evoked potentials to determine hearing thresholds. CABRA includes an expert system to assist users in identifying auditory responses, while providng a user-friendly interface for test configuration and result visualization. Its portability, affordability, and focus on accessibility make it a promising tool for screening hearing loss in infants and patients unable to respond actively. Future improvements include hardware latency reduction, integration of machine learning for enhanced signal interpretation, and a more compact design.