Assess the ability to design an original annotation schema and an end-to-end computational framework to identify, model, and generate automated feedback on discourse-level rhetorical moves in multilingual doctoral theses across disciplines. Tasks include: specifying annotation categories and guidelines with inter-annotator reliability plans; ensuring cross-linguistic and cross-disciplinary validity; outlining modeling approaches (representations, sequence/classification models, evaluation metrics); designing feedback generation, user interaction, and ethical/deployment considerations. Provide justified design choices and sample annotated excerpts.