Robotic rehabilitation technologies have become increasingly important in addressing functional mobility limitations, affecting millions of people globally. These technologies offer personalized therapy through consistent and precise control of limb movements. User feedback is important to improve these systems´ design and functionality and attain better rehabilitation outcomes. The Brazo-FiT system is a rehabilitation tool designed for upper extremity therapy, operated through a brain-computer interface (BCI) and comprising a UR3 collaborative robot and a supportive setup including a height-adjustable chair, a large display, and an Arduino controller. This setup enables personalized therapy by interpreting users’ imagined movements and translating them into physical arm and forearm movements via the robot. This study evaluates Brazo-FiT performance and user experience. Brazo-FiT demonstrates effective interpretation of movement imagination with an average training phase classification accuracy of 90% and 76% after cross-validation and a 60% success rate during online testing. User satisfaction is notably high, with positive feedback on the system’s efficiency, appearance, smoothness of operation, and component comfort. Open-ended responses suggest areas for improvement, including more personalized session pacing and enhanced cognitive guidance for movement imagination.