Quality requirements for additive manufacturing processes are rising as these technologies are increasingly used for high-quality and critical components in various industries, such as aerospace, automotive, medical technology, and biotechnology. Due to its design flexibility, additive manufacturing is often used for small and custom series, making quality control a challenge. A promising approach to improve cost-effectiveness is process monitoring using machine learning. This study presents a sensor concept for fused deposition modeling and a machine learning pipeline to predict process and part quality. Tensile and impact strength were selected as key quality characteristics. The results show high accuracy in predicting tensile strength, while lower accuracy for impact strength is attributed to data variability caused by natural scattering in samples with identical process conditions. This highlights machine learning’s potential to enhance efficiency and reduce costs, offering an alternative to expensive testing methods.