Doctoral education faces high attrition rates and widespread well-being challenges. Research has identified key motivational factors distinguishing persisting students from those who withdraw: perceived progress, exhaustion levels, and thesis topic appropriation. Despite this knowledge, technological interventions, especially AI-based ones, targeting doctoral education remain scarce, partly due to each doctoral journey’s uniqueness. This paper presents the co-design process and initial evaluations of a novel generative AI (GenAI)-enhanced system supporting doctoral students’ perception of progress and well-being. The system facilitates practices (e.g., journaling, reflection) targeting the three aforementioned motivational factors that doctoral training interventions suggest can to improve student well-being. The system’s key innovation is its combination of machine learning-based open learner models (based on individual’s journaling) with GenAI explanations, enabling personalized student support for the analysis, visualization and sense-making of their progress and their specific predictive factors. Initial evaluations with full-time (N = 8) and part-time (N = 10) doctoral student cohorts during 4-week training periods show high usability and likelihood-to-recommend. Feedback also reveals improvement areas and design tensions, particularly regarding strategies for consistent platform engagement.