In this demo, we generate an ontology directly from your data, no manual modeling required.The Assistant doesn't guess. It explores your actual data to build a graph schema you can trust:

• Surveys your catalog: which tables exist, what columns they have, and their types
• Samples real rows and profiles columns to spot IDs, enums, free text, and timestamps
• Probes foreign-key relationships by testing joins between candidate tables against real data and measuring the hit rate
• Drafts a graph schema, deciding which tables become vertices, which become edges, and what the key columns are
• Proposes each vertex and edge to you one at a time for review
• Verifies that each label returns rows once applied

The result is an AI-ready semantic model of your data, exportable in Apache Ossie format.