Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Design Science Blueprint for an Orchestrated AI Assistant in Doctoral Supervision

Teo Susnjak, Timothy R. McIntosh, Tong Liu, Paul Watters

TL;DR

The paper tackles the problem of doctoral supervision strain amid rapid AI adoption and rising online enrollment by proposing a design science blueprint for an orchestrated AI assistant that mediates between students, supervisors, and graduate research schools. It grounds the artefact in Stakeholder Theory and Academic Integrity, mapping supervision pain points to a stack that includes retrieval-augmented generation, temporal knowledge graphs, and mixture-of-experts routing, all anchored by a student context store and behavior patches. The main contributions are a theoretically informed mediation framework, a practical, executable workflow with governance rules, and an ex ante risk assessment that highlights where automation can reduce latency while preserving human judgment. The work is intended as a template institutions can trial across disciplines, with emphasis on transparency, auditable decision-making, and safeguarding against the illusion of learning.

Abstract

This study presents a design science blueprint for an orchestrated AI assistant and co-pilot in doctoral supervision that acts as a socio-technical mediator. Design requirements are derived from Stakeholder Theory and bounded by Academic Integrity. We consolidated recent evidence on supervision gaps and student wellbeing, then mapped issues to adjacent large language model capabilities using a transparent severity-mitigability triage. The artefact assembles existing capabilities into one accountable agentic AI workflow that proposes retrieval-augmented generation and temporal knowledge graphs, as well as mixture-of-experts routing as a solution stack of technologies to address existing doctoral supervision pain points. Additionally, a student context store is proposed, which introduces behaviour patches that turn tacit guidance into auditable practice and student-set thresholds that trigger progress summaries, while keeping authorship and final judgement with people. We specify a student-initiated moderation loop in which assistant outputs are routed to a supervisor for review and patching, and we analyse a reconfigured stakeholder ecosystem that makes information explicit and accountable. Risks in such a system exist, and among others, include AI over-reliance and the potential for the illusion of learning, while guardrails are proposed. The contribution is an ex ante, literature-grounded design with workflow and governance rules that institutions can implement and trial across disciplines.

A Design Science Blueprint for an Orchestrated AI Assistant in Doctoral Supervision

TL;DR

The paper tackles the problem of doctoral supervision strain amid rapid AI adoption and rising online enrollment by proposing a design science blueprint for an orchestrated AI assistant that mediates between students, supervisors, and graduate research schools. It grounds the artefact in Stakeholder Theory and Academic Integrity, mapping supervision pain points to a stack that includes retrieval-augmented generation, temporal knowledge graphs, and mixture-of-experts routing, all anchored by a student context store and behavior patches. The main contributions are a theoretically informed mediation framework, a practical, executable workflow with governance rules, and an ex ante risk assessment that highlights where automation can reduce latency while preserving human judgment. The work is intended as a template institutions can trial across disciplines, with emphasis on transparency, auditable decision-making, and safeguarding against the illusion of learning.

Abstract

This study presents a design science blueprint for an orchestrated AI assistant and co-pilot in doctoral supervision that acts as a socio-technical mediator. Design requirements are derived from Stakeholder Theory and bounded by Academic Integrity. We consolidated recent evidence on supervision gaps and student wellbeing, then mapped issues to adjacent large language model capabilities using a transparent severity-mitigability triage. The artefact assembles existing capabilities into one accountable agentic AI workflow that proposes retrieval-augmented generation and temporal knowledge graphs, as well as mixture-of-experts routing as a solution stack of technologies to address existing doctoral supervision pain points. Additionally, a student context store is proposed, which introduces behaviour patches that turn tacit guidance into auditable practice and student-set thresholds that trigger progress summaries, while keeping authorship and final judgement with people. We specify a student-initiated moderation loop in which assistant outputs are routed to a supervisor for review and patching, and we analyse a reconfigured stakeholder ecosystem that makes information explicit and accountable. Risks in such a system exist, and among others, include AI over-reliance and the potential for the illusion of learning, while guardrails are proposed. The contribution is an ex ante, literature-grounded design with workflow and governance rules that institutions can implement and trial across disciplines.
Paper Structure (54 sections, 1 equation, 4 figures, 3 tables)

This paper contains 54 sections, 1 equation, 4 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Research Aims and Objectives
  3. Scope and Significance of the Study
  4. Literature Review
  5. PhD Student's Identity Formation
  6. Typical Task Distribution in PhD Supervision
  7. Gaps and Inefficiencies in Human PhD Supervisions
  8. LLMs and Their Emerging Capabilities
  9. How these parts fit together.
  10. Mapping the Orchestrated System to Bloom's Taxonomy
  11. Research Methodology: A Design Science Approach
  12. Methods
  13. Phase 1: Literature grounding.
  14. Phase 2: Artefact design and justification.
  15. Ex ante criteria application.
  16. ...and 39 more sections

Figures (4)

  • Figure 1: Estimates of typical PhD supervision tasks and effort distribution bands AitchisonLee2006Granello2004Bogelund2015levecque2017work.
  • Figure 2: Stakeholders and their relationships in the context of PhD supervision
  • Figure 3: A proposed workflow segment for AI-assisted PhD supervision. The diagram depicts the student–AI loop and the student-initiated moderation loop with the human supervisor, where supervisor patches persist and steer later responses. Students can set goals and thresholds that drive optional progress summaries.
  • Figure 4: Reconfigured stakeholder ecosystem with the AI assistant acting as a socio-technical mediator.