LegalFab uses AI-driven ontology generation rather than predefined legal concepts - the system is approximately 90% automated, with 10% requiring human review and validation. The legal concepts automatically extracted from your ingested documents - historical cases, statutes, and legal sources - build a jurisdiction-specific ontology tailored to UK or US law. The system naturally accounts for jurisdictional differences (e.g., "claimant" vs. "plaintiff," different damage calculations) by learning from the source materials you provide. Your legal team reviews the auto-generated ontology through the Galaxy interface during initial setup to validate critical legal concepts, refine edge cases, and ensure domain accuracy, but the heavy lifting of concept extraction, relationship mapping, and entity classification is handled automatically. This 90/10 approach enables rapid deployment across multiple jurisdictions (weeks rather than months) while maintaining the legal precision that requires expert judgment, without the burden of manually constructing the entire knowledge framework.