Pipeline Dashboard
Phase 1
complete
PDF Ingestion
Extract document structure from regulatory PDF using vision model.
Phase 2
complete
Reference Extraction
Identify internal and external cross-references between sections.
Phase 3
complete
Rule Generation
Generate Turnip DSL rules encoding legal obligations and permissions.
Phase 4
complete
Graph Building
Build the Realta property graph from parsed Turnip rules.
Phase 5
pending
Reasoner
Query the graph with user-supplied facts to derive compliance conclusions.