Universities increasingly face the challenge of supporting diverse engineering students through critical transitions, while leveraging mentoring and digital infrastructures as strategic levers for knowledge management and student success. Building on mentoring and communities‐of‐practice literatures, this paper conceptualises mentoring in an engineering department as a sociodigital ecosystem that integrates formal, informal and peer‐based practices with digital communities and AI‐enabled tools. Empirically, the study focuses on the “continuous mentoring” initiative developed at the Engineering Department of a medium‐sized private university in Southern Italy, where a department‐wide digital community, cohort and event‐based groups, course‐specific chats and project teams have progressively evolved into a multi‐layered knowledge infrastructure.
The research adopts a qualitative single‐case design, drawing on internal project reports, departmental documents, social media analytics, extended abstracts and digital traces from student communities. These materials were coded inductively and organised into an ecosystem representation capturing layers, roles and knowledge flows, then interpreted through the lens of mentoring ecosystems and socio‐digital knowledge infrastructures. The findings show how a persistent core community, surrounded by cohort‐specific and project‐based groups, supports first‐year students by lowering the threshold for asking questions, sharing tacit tips and accessing timely guidance from faculty mentors, peer tutors and community administrators. The same ecosystem also amplifies participation in extracurricular and research‐oriented initiatives, such as the European Researchers Night and the Ingenium entrepreneurship team, enabling students to mobilise situated know‐how on how to prepare, present and learn from comparison with peers from other institutions.
Theoretically, the paper extends mentoring research by framing mentoring not as a single programme, but as an ecosystem of interconnected practices embedded in socio‐digital infrastructures, where human relations, community practices and AI‐enabled assistants jointly sustain knowledge creation and sharing. Practically, it identifies three design principles for universities seeking to build mentoring‐based knowledge infrastructures: anchoring mentoring in a persistent core community; cultivating flexible satellite communities around cohorts, courses and projects; and augmenting, rather than replacing, human mentoring with AI tools for routine queries. The contribution speaks to scholars and practitioners interested in mentoring, knowledge management and the design of learning ecosystems in engineering and beyond.