Chen Cheng ¹ and Moddwyn Andaya ² , ¹ USA, ² California State University, USA

Many students struggle to remain motivated when practicing art independently due to limited access to personalized guidance and feedback. This paper presents ArtNavi, an AI-powered art tutoring application designed to provide step-by-step instruction, real-time feedback, and motivational support. The system integrates Unity and C# with OpenAI services, combining a modular manager architecture, conversational AI, and persistent project storage to create an interactive learning environment [7]. Key implementation challenges included maintaining responsive asynchronous communication, ensuring reliable data persistence, and preserving consistent AI feedback quality; these were addressed through structured request handling, modular data serialization, and guarded message pipelines. Experimental evaluation examined feedback effectiveness across beginner and intermediate scenarios. Results showed strong performance for novice guidance, with slightly reduced effectiveness for more advanced prompts due to increased critique complexity. Overall, ArtNavi demonstrates that AI-driven tutoring can make creative learning more accessible, scalable, and supportive for diverse learners

AI Tutoring, Art Education, Interactive Learning, Creative Feedback