We defined a comprehensive information model to describe all entities and relationships of predictive workflows:
From business context to intermediate stages and trained models, up to computed predictions.
We use our connectivity model to enrich each stage of a predictive workflow with comprehensive meta information.
Enrichment is key to understand every workflow and its generated result.
We offer a purpose-built plugin to organize enriched predictive workflows and their results as information networks.
These networks are sent to a graph database of choice to get connected to all other enriched workflows.