iLeapTutor: a RAG-enabled generative AI platform for standards-aligned university course authoring

Introduction Recent advances in Artificial intelligence (AI) and Generative AI (GenAI) have enabled more dynamic digital learning experiences. Intelligent Tutoring Systems (ITSs) leverage this capability to improve learning outcomes by offering structured and responsive support. This paper introduces iLeap Tutor , an intelligent learning platform that delivers consistent, structured, course-aligned learning materials that meet course requirements, such as Course Learning Outcomes (CLOs) and National E-learning Center (NeLC) standards. Methods iLeap Tutor uses retrieval-augmented generation (RAG) over the tutor's uploaded course materials together with Gemini 3.1 Flash-Lite (Preview), with prompts that embed the CLO and NeLC requirements, and the tutor reviews and approves the generated course. The platform was evaluated as a single-course proof-of-concept using the Selected Topics in Computer Science course at Imam Abdulrahman Bin Faisal University (IAU). The generated content was assessed using BERTScore, Self-BLEU, and latency measurements, and thirteen academic experts evaluated the generated course in terms of content accuracy, alignment with CLOs, structure, language quality, and platform usability. Results The evaluation achieved a BERTScore-F1 of 0.829 and a cross-run Self-BLEU result of 0.553 and a within-lesson Self-BLEU of 0.146, while the generation of 16 course topics averaged 219 s in latency. At least 92.3% of responses on every questionnaire item indicated agreement, and all of them would recommend the platform. Discussion The results indicate strong semantic alignment and output consistency across multiple runs and within-lesson content diversity.