Operationalising Consulting Methodology Through Workflow Orchestration: A Design Science Study of CWORT with Governed Agentic AI Support

Abstract

Consulting engagements play a critical role in organisational decision-making, yet delivery often remains highly manual, fragmented, and dependent on individual practitioner experience. While consulting methodologies provide conceptual guidance, they are rarely operationalised as executable systems, leading to variability in delivery quality, limited traceability between inputs and outputs, and challenges in scaling consistent advisory practices. The growing use of digital and artificial intelligence (AI)-enabled tools in consulting has tended to focus on isolated tasks rather than the orchestration of end-to-end workflows, raising concerns around governance and accountability. This paper introduces CWORT (Consulting Workflow Orchestration Tool), a socio-technical system that operationalises consulting methodology through a digitally governed, workflow-based approach. CWORT represents consulting delivery as a sequence of explicit workflow states, integrating role-based governance, artefact traceability, and constrained agentic reasoning components to support analysis and synthesis activities. AI capabilities are embedded within predefined workflow stages and operate under strict human-in-the-loop control, ensuring that professional judgement and accountability remain central to advisory outcomes. The study is framed within a design science research paradigm and presents the conceptual model, system architecture, and a structured case-based evaluation in a real consulting engagement. The evaluation adopts a within-context comparative field design, demonstrating that workflow orchestration improves transparency, consistency, and traceability of consulting outputs, while enabling substantial reductions in discovery and assessment effort relative to prior approaches. AI-supported components are shown to augment interpretation and insight generation without altering deterministic analytical outcomes.

Received: 2 January 2026 | Revised: 7 April 2026 | Accepted: 12 May 2026 

Conflicts of Interest

The author declares that he has no conflicts of interest to this work.

Data Availability Statement

Data sharing is not applicable to this article as no new data were created or analysed in this study.

Author Contribution Statement

Oluwaseun Iyiola: Conceptualisation, Methodology, Software, Writing – original draft, Writing – review & editing, Visualisation, Supervision.