Evidence · first-party tested

The tool derives semantic transaction labels instead of leaving raw descriptions only; it classifies entries with transaction_type values such as UPI and transaction_id values such as 917615251879.

✓ WorkedRetab
Input — the actual file we sent
Bank Statement PDF
A 4-page bank statement PDF with 51 transaction rows, balance-forward continuations, account/branch metadata, rewards information, and disclaimer text. Designed to test end-to-end structured extraction from a dense financial statement.
Output — unretouched
tool output
The observation

The tool derives semantic transaction labels instead of leaving raw descriptions only; it classifies entries with transaction_type values such as UPI and transaction_id values such as 917615251879.

Criterion: Semantic Field Enrichment · Scenario: Bank Statement PDF (group: financial-document-extraction)
Query this
get_evidence({
  tool: "retab",
  scenario: "financial-document-extraction"
})
MCP · mcp.aidemos.com/api/mcp
Free with attribution.
Real outputs, no retouching · every cell queryable via API & MCP · aidemos.com