/home/michael

Higher Education Reference Model (HERM)

3 min read

Note

Why this matters

When AI agents consume institutional documentation as operational input, they need typed, explicit relationships between entities — not narrative prose. HERM provides a standardised vocabulary for those relationships across the higher education sector: courses relate to programmes, programmes lead to degrees, degrees map to competencies. It is not the only possible ontology, but it is a pragmatic and widely adopted starting point for making institutional knowledge machine-readable.

Higher Education Reference Model (HERM)

One-sentence definition: HERM is a standardised set of reference architectures — business, data, and application — that provides a common vocabulary for describing how higher education institutions are structured and how their processes, systems, and data relate to each other.

HERM is maintained collaboratively by CAUDIT (Council of Australasian University Directors of Information Technology) and UCISA (Universities and Colleges Information Systems Association) and is currently at version 3.2.0 (CAUDIT, 2025; UCISA, 2025). It is used by institutions internationally as a foundation for enterprise architecture, systems integration, and — increasingly — for structuring institutional data in ways that support AI-mediated processes.

What it provides

HERM defines entity types (programmes, modules, assessments, competencies, students, staff) and the typed relationships between them. A programme contains modules. A module has learning outcomes. An assessment tests outcomes. Outcomes map to competencies. These relationships are explicit and directional — precisely the kind of structure that AI agents require for reliable operation.

The model operates at institutional scale, addressing how major functional areas (curriculum, research, student administration, finance) relate to one another and to the data and systems that support them. It does not prescribe how individual teams or educators should organise their working knowledge, but it provides the overarching vocabulary within which more granular structures can sit.

Where it fits in the infrastructure argument

In the context of documentation as infrastructure, HERM represents what institutions already have — or could have — at the structural level. The problem is that HERM-level relationships are typically documented in narrative prose (module specifications, programme handbooks) rather than encoded as queryable data. The graph infrastructure proposal argues for making these relationships computationally traversable rather than leaving them embedded in documents that require human interpretation.

HERM provides a pragmatic starting point for ontology design, but it is not the only option. Institutions may need to extend or adapt it to represent relationships that HERM does not capture well — for instance, the distinction between a clinical placement and a practice competency in nursing education, or the cross-disciplinary connections in an integrated medical curriculum. What matters is internal consistency rather than strict adherence to an external standard. Graph databases are schema-flexible, and ontologies can evolve as institutional needs develop.

What it does not address

HERM operates at the level of institutional architecture. It does not address how individuals or teams structure the operational knowledge that sits beneath and between institutional systems — the project documentation, meeting records, and professional workflows that constitute the day-to-day fabric of institutional life. That gap is where the documentation debt problem is most acute and where the argument for structured operational architecture is most pressing.

Sources

Notes

Referenced in both documentation-as-infrastructure and curriculum-infrastructure, where it provides the ontological foundation for the three-layer architecture proposal (graph database + vector database + Model Context Protocol).

Continue reading

  • Documentation becomes infrastructure when AI agents are the readers

    When AI agents consume documentation as operational input, it undergoes a category shift from reference material to operational architecture — inaccuracies no longer merely inconvenience readers, they cause system failures. This essay argues that the primary bottleneck for institutional AI integration is not AI capability but information architecture: how institutional knowledge is structured, maintained, and made available to AI systems. Documentation written for human readers cannot function as reliable AI input without deliberate restructuring around explicit relationships and rigorous maintenance workflows. Treating this transition as a governance imperative — rather than a technical afterthought — determines whether AI integration delivers on its institutional promise.

  • Beyond document management: Graph infrastructure for professional education curricula

    Professional curricula are extensively documented but not systematically queryable, creating artificial information scarcity that makes compliance reporting and quality assurance labour-intensive. This essay proposes a three-layer architecture — graph databases as the source of truth for curriculum structure, vector databases for semantic content retrieval, and a Model Context Protocol layer for stakeholder access — that transforms documentation into operational infrastructure. The architecture incorporates temporal versioning for longitudinal evidence, role-based access controls for multi-stakeholder environments, and internal quality audit against institutional policy alongside external regulatory compliance, enabling verification in hours rather than weeks.

  • Documentation debt

    The accumulated cost of outdated, ambiguous, or poorly structured institutional knowledge — manageable when humans compensate, operationally consequential when AI agents depend on it literally.

Graph View

Backlinks