Read case study: Teaching VETR Which Clauses Actually BindTeaching VETR Which Clauses Actually BindVETR Proposal (Internal R&D)+0.08–0.14 F1Binding-set accuracy gain over regexes70–79%Specificity vs 8–19% at equal recallRead case study
Read case study: Building the first open benchmark for federal contracting AI: how we validated AI-assisted FAR clause extractionFederal Contracting / Government ProcurementBuilding the first open benchmark for federal contracting AI: how we validated AI-assisted FAR clause extractionVETR Proposal (Internal R&D)0%Claude Sonnet 4.6 hallucination rate on real FAR text32.1%Claude Haiku 4.5 hallucination rate on real FAR textRead case study