.png){kind=icon}
.png){kind=icon}
Return
Beyond Assistance: Orchestrating the Developmental Triad for Agentic AI Mentorship
.png){kind=icon}
PRE
AI
W
C
D
DOI：10.1080/10447318.2026.2687899.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Junior UX designers face a significant mentorship gap, as current AI tools focus on discrete tasks rather than holistic growth. We propose the Developmental Triad, a multi-persona AI model with three agentic functions: an Expert (William) for guidance, a Companion (Maya) for psychosocial support, and an Assistant (Cody) for executional aid. Informed by interviews with 20 experts, we evaluated this model through a four-week study with eight junior designers, followed by in-depth reflective interviews to capture the nuances of their interaction. Findings reveal a “Ping-Pong” interaction pattern, where mentees strategically navigate between personas to manage diverse cognitive and emotional demands. The system’s social judgment-free environment was perceived to enhance professional self-efficacy and supported the formation of professional mental models, preparing mentees for real-world workplace demands. This research establishes a scalable framework for agentic AI mentorship through Meta-Orchestration, ultimately democratizing professional expertise in creative industries.
Keywords:
Agentic AI
AI-mediated mentorship
Developmental Triad
meta-orchestration
strategic friction
Journal
I
IF:
0
Papers:
329
Citations:
0