Corrections

When the extraction pipeline produces incorrect values, Talonic provides both manual and automated correction workflows. Edit values directly in the structured data grid, apply bulk corrections using rules, or flag cells for review. Every correction is tracked in the provenance trail, and corrections feed back into the pipeline as training signals to improve future extraction accuracy for similar documents.