The challenge
Two departments worked on the same incidents without speaking the same language: different terms, reference data and tools. The incident history existed, but it was hard to use it to understand causes and prevent incidents from happening again.
The value delivered
DEEPXP modelled a shared ontology in Protégé that maps both departments’ vocabularies, then built a knowledge graph centred on the incident history. A machine-learning similarity search finds the past incidents closest to a new one, with their root cause and the solution that was applied. A continuous improvement loop collects feedback from domain experts to enrich the knowledge in the graph.
Results
Both departments now share one view of incidents. When a new incident occurs, teams find similar cases, their likely root cause and the solutions already applied, instead of starting from scratch.