SOP Library PoC

When I joined Eurotux, all its technical documentation lived in an XWiki self-hosted implementation. It worked well for collaborative authoring, but for procedural pieces it had a structural weakness: there wasn’t a simple way to enforce consistent formatting and metadata, nor to implement a validation workflow, things to which a valid SOP must comply. A technician could omit critical metadata (objective, environment, tools, authorisations) and the system would accept it. An approved SOP could be silently modified without triggering a new review cycle.

After I completed an XML certification course, I designed and built a proof of concept that stored each SOP as an XML file governed by an XML Schema that declared exactly which fields a valid procedure must contain, rejecting documents that failed validation.

Migrating from XWiki to an XML architecture would also serve a second goal: improve LLM ingestion. Structured XML is more reliably parseable by AI than XWiki markup, thus making the library AI-ready for future automated workflows.

The web interface is visually simple, but it illustrates the intended features: i) browsing the library and ii) viewing, editing, and printing SOPs. Additionally, it also proves the concept of a rule-enforced architecture, because behind every edit, schema validation ran and blocked invalid edits. In production, every save would trigger a pull request.

The proof of concept was well received, but full implementation was deferred pending RBAC support from the internal tools team.