Review Result

Reclassify individual match results from a matching run as matched, review, or no_match. Matched results feed downstream enrichment and reconciliation.

Review a single match result by reclassifying its status. Each result carries the status the engine assigned when the run executed — matched, review, or no_match — and this endpoint moves it between those same three states after human inspection. Set matched to accept a borderline candidate, no_match to reject a false positive, or review to flag a result for another look.

The typical review workflow targets the review band: results whose confidence landed between the review floor (0.4) and the config threshold. Inspect the evidence.field_contributions to see exactly which fields agreed and disagreed, then accept (matched) or reject (no_match). Reclassifying does not recompute confidence — the score remains the engine's original weighted sum.

The response returns the full updated match result object, including the evidence breakdown and the new status. Once reviewed, a result can be re-reviewed by calling this endpoint again with a different status — there is no lock on the review state. Any other status value is rejected with a 400 listing the valid values.

For bulk review workflows, iterate over results from GET /v1/matching/runs/:id/results?status=review and call this endpoint for each result. The API enforces standard rate limits — for high-volume batch review, use reasonable concurrency (e.g. 10 concurrent requests) and respect the Retry-After header on 429 responses.

{
  "status": "matched"
}

Response

{
  "id": "e5f6a7b8-c9d0-1234-efab-345678901234",
  "document_id": "doc_uuid_1",
  "document_filename": "invoice-acme-2024-001.pdf",
  "matched_reference_row_id": "ref_row_42",
  "confidence": 0.78,
  "status": "matched",
  "evidence": {
    "field_contributions": [
      { "extracted_field": "vendor_name", "extracted_value": "Acme Corp GmbH", "reference_value": "Acme Corp", "match_type": "fuzzy_string", "matched": true, "weight": 0.4, "score": 0.95, "contribution": 0.38 },
      { "extracted_field": "invoice_date", "extracted_value": "2024-09-28", "reference_value": "2024-09-30", "match_type": "date_range", "matched": true, "weight": 0.3, "score": 1.0, "contribution": 0.3 },
      { "extracted_field": "amount", "extracted_value": 14100.00, "reference_value": 14250.50, "match_type": "numeric_range", "matched": false, "weight": 0.3, "score": 0.33, "contribution": 0.1 }
    ],
    "candidates_considered": 3,
    "top_n_candidates": [
      { "reference_row_id": "ref_row_42", "confidence": 0.78 }
    ],
    "blocking_method": null
  }
}

Errors