A Replicable Low-Code CRM Reference Architecture for Small Higher-Education Administrative Units

Abstract

Small administrative units in higher education—continuing-education offices, language and culture centers, advising and tutoring services, and similar back-office functions—are routinely expected to deliver responsive, data-informed service with the budgets and staffing of a support desk. Enterprise customer relationship management (CRM) platforms are frequently beyond their financial and administrative reach, and the common fallback of spreadsheets, shared inboxes, and manual tracking does not scale. This paper presents a replicable reference architecture for building a custom CRM on the Microsoft Power Platform—Power Apps, Power Automate, SQL Server, SharePoint, and Power BI—assembled and maintained by a single technologist rather than a dedicated development team. Drawing on the author's experience developing such a system for a university continuing-education unit, the paper generalizes that experience into a transferable set of architectural decisions, design patterns, and connector-based integration mechanisms, and it argues that the low-code paradigm makes enterprise-grade administrative software attainable for units that could never commission bespoke development. The paper further sets out a forward-looking framework for augmenting this architecture with artificial intelligence and large language models—intake triage, document processing, drafting and summarization, and conversational self-service—implemented through tools such as Copilot Studio and AI Builder and grounded in the current research literature. The AI layer is presented explicitly as a proposed extension and research agenda rather than a deployed result, and it is accompanied by a treatment of the governance, data-privacy, and human-oversight guardrails that responsible adoption requires. The contribution is a practical blueprint that small units can evaluate against their own context and adopt incrementally.