The article presents the concept, implementation, and effectiveness assessment of a vision system incorporating elements of artificial intelligence (AI) to support quality control in the packaging process of orthopedic bone wax. The research problem addressed the inherent limitations of human visual inspection -namely, variability in assessment outcomes, cognitive load imposed on operators, and difficulties in detecting defects on transparent or glossy surfaces - as well as the need to ensure repeatable and objective real-time inspection. A system was designed that leverages AI-based tools in the learning process of the inspection program. The effectiveness evaluation, conducted under real production conditions, demonstrated a reduction in the number of nonconformities identified during the final inspection (after the packaging operation), accompanied by a decrease in cycle time and improved process stability. The proposed approach constitutes a practical example of AI implementation in the medical device sector, emphasizing product compliance with regulatory requirements and system robustness against process disturbances.