Upload two versions of the same Barnevernet document — or any two related documents — and the agent will find contradictions, deleted facts, new allegations, and party changes between them. Results include corpus-backed legal significance for each discrepancy.

Document A — Earlier / Original

Drop here or

PDF, DOCX, TXT — max 8 MB

Document B — Later / Comparison

Drop here or

PDF, DOCX, TXT — max 8 MB

Engine

Engine applies to the final synthesis only. Norwegian specialist v3 excels at identifying legally significant discrepancies in Barnevernet documents — procedural violations, threshold errors, and missing statutory justifications. Classification, party extraction, timelines, and cross-referencing always use azure-mini.

Corpus slices (used for legal significance context)

The corpus provides legal significance context for each discrepancy found. All four default slices cover the core Barnevernet framework.

Ready

Upload two Barnevernet documents, then run. The agent will classify each document, extract parties and timelines, cross-reference them for discrepancies, and produce a corpus-backed legal significance report.

Typical use: compare the original Bekymringsmelding against the later Vedtak, or compare two versions of a Barnevernet investigation report.