We present work in progress on a score-informed automatic music transcription and performance assessment system designed to support automated musical education. The system combines transcription models with score alignment and error detection tools to provide meaningful feedback to both students and teachers. It comprises an automatic guitar transcription model trained using a combination of public and internal data, along with pre-trained transcription models for piano and violin. Following transcription, the system evaluates performances by aligning the transcribed results with reference music scores to identify various errors, such as pitch, rhythm, tempo, structural, and intonation. This work demonstrates how intelligent systems can enhance the music learning experience by making it more interactive, personalized, and data-driven.