{"observation":{"id":"010e7dbd-9574-4e0f-bdc3-c370101d3914","tool":"extend-ai","tool_name":"Extend AI","criterion":"structural-clean-output","criterion_name":"Structural Clean Output","criterion_definition":"Is the JSON directly consumable by a downstream AI pipeline or system without requiring a structural transformation layer?","criterion_evidence_type":"transformation","criterion_rank_role":"context","criterion_rank_role_reason":"Being directly consumable by a downstream pipeline is valuable, but it is more about integration convenience than whether the tool actually extracts the data correctly. (3 of 3 judges)","scenario":"bank-statement-pdf","scenario_name":"Bank Statement PDF","group_tag":"business-document-extraction","scenario_description":"A four-page bank statement PDF with dense transaction tables, balance-forward bridges, account metadata, rewards data, and disclaimer text. It was used to stress schema-driven extraction, multi-page continuity, row completeness, and financial numerical accuracy.","modality":"pdf","input_text":null,"input_artifact_refs":[{"alt":null,"url":"https://cdn.futuresmart.ai/public/aidemos/a459a9de9c8a466d8f69b112e04ddad2.pdf?v=1","role":"input","filename":"Bank Statement PDF.pdf"}],"stresses":["Table extraction across 50+ transaction rows","Multi-page continuity with balance-forward bridges","Parsing structured account metadata alongside unstructured transaction descriptions","Numerical accuracy for deposits, withdrawals, running balances, and summaries","Extraction of nested rewards and disclaimer sections"],"verdict":"failed","score":null,"score_total":null,"note":"The output reorders top-level schema objects instead of preserving the declared sequence, so consumers that depend on the original order need an extra transformation step.","evidence_state":"verified","source":null,"artifacts":[{"url":"https://d3epheqghktydj.cloudfront.net/research-media-extend-ai-bank-statement-schema-order-69fcdddefcea.png","role":"output","alt":null}],"run_id":"a061b9e7-a9c5-443d-a171-b296aaf51b8c","study_title":"Extract and query structured data from documents using natural language","study_kind":"generation","research_task":"86b9y25e5","tested_at":null,"completeness":"input-and-output","input":{"state":"files","text":null,"files":[{"url":"https://cdn.futuresmart.ai/public/aidemos/a459a9de9c8a466d8f69b112e04ddad2.pdf?v=1","filename":"Bank Statement PDF.pdf","alt":"Bank Statement PDF","role":"input"}],"modality":"pdf","stresses":["Table extraction across 50+ transaction rows","Multi-page continuity with balance-forward bridges","Parsing structured account metadata alongside unstructured transaction descriptions","Numerical accuracy for deposits, withdrawals, running balances, and summaries","Extraction of nested rewards and disclaimer sections"]},"tool_page_slug":"extend-ai","tool_url":"https://aidemos.com/tools/extend-ai","permalink":"https://aidemos.com/evidence/010e7dbd-9574-4e0f-bdc3-c370101d3914","api_url":"https://ai.aidemos.com/v1/observations/010e7dbd-9574-4e0f-bdc3-c370101d3914"},"peers":[{"id":"6d4aa957-2bb9-4651-9de3-19ae97c912f3","tool":"landing-ai","tool_name":"Landing AI","verdict":"worked","score":null,"score_total":null,"note":"Delivers JSON that is directly usable downstream, with the demo moving from upload to extraction results and the report noting downloadable output with no extra transformation step.","artifact_count":2,"thumbnail":"https://cdn.futuresmart.ai/public/aidemos/48962ee031ec4496b0c08a3aacebe0d8.mp4?v=1","evidence_url":"https://aidemos.com/evidence/6d4aa957-2bb9-4651-9de3-19ae97c912f3"},{"id":"527704b6-b5a4-4c3d-ba5e-e7d4275464b9","tool":"retab","tool_name":"Retab","verdict":"worked","score":null,"score_total":null,"note":"Produces directly copyable JSON from the workflow, so the bank-statement extraction is immediately usable without a transformation layer after configuration.","artifact_count":3,"thumbnail":"https://d3epheqghktydj.cloudfront.net/research-media-bank-statement-2-jul-a511cf7bd0d6.png","evidence_url":"https://aidemos.com/evidence/527704b6-b5a4-4c3d-ba5e-e7d4275464b9"}],"other_criteria":[{"id":"310a731e-d9ba-4ab5-9782-fb4b8d4dabdb","criterion":"extraction-accuracy","criterion_name":"Extraction Accuracy","rank_role":"decisive","verdict":"failed","score":null,"score_total":null,"note":"The summary aggregation is incorrect: `total_transactions` is 49 in the output, while the report says the expected count after exclusions is 40.","artifact_count":2,"evidence_url":"https://aidemos.com/evidence/310a731e-d9ba-4ab5-9782-fb4b8d4dabdb"},{"id":"b1ccd11a-19d7-4af9-93c2-1ba1086296e5","criterion":"schema-adherence","criterion_name":"Schema Adherence","rank_role":"decisive","verdict":"worked","score":null,"score_total":null,"note":"The tool reconstructs the bank-statement hierarchy into nested JSON with branch, account, rewards, metadata, balances, summary, and transaction-related objects present in the requested layout.","artifact_count":5,"evidence_url":"https://aidemos.com/evidence/b1ccd11a-19d7-4af9-93c2-1ba1086296e5"},{"id":"ca67652c-8a96-456d-92ef-c2817c381589","criterion":"semantic-field-enrichment","criterion_name":"Semantic Field Enrichment","rank_role":"decisive","verdict":"worked","score":null,"score_total":null,"note":"The tool infers transaction direction from free text, classifying an ATM entry as `Withdrawal` and populating `withdrawal_amount: 1000`.","artifact_count":2,"evidence_url":"https://aidemos.com/evidence/ca67652c-8a96-456d-92ef-c2817c381589"},{"id":"2c08bc5e-efaf-45c1-91d9-08a43c075910","criterion":"semantic-field-enrichment","criterion_name":"Semantic Field Enrichment","rank_role":"decisive","verdict":"failed","score":null,"score_total":null,"note":"It leaves `transaction_id` null even when the description contains embedded reference numbers, so identifier extraction is not recovered from the transaction text.","artifact_count":1,"evidence_url":"https://aidemos.com/evidence/2c08bc5e-efaf-45c1-91d9-08a43c075910"}],"appears_in":[{"page_type":"ranking","slug":"document-extraction","title":"Best AI Tools for Extracting Structured Data from Business Documents","url":"https://aidemos.com/best/document-extraction","binding":"run"},{"page_type":"tool","slug":"docsumo","title":null,"url":"https://aidemos.com/tools/docsumo","binding":"run"}],"same_scenario":[{"id":"6d4aa957-2bb9-4651-9de3-19ae97c912f3","tool":"landing-ai","tool_name":"Landing AI","verdict":"worked","score":null,"score_total":null,"note":"Delivers JSON that is directly usable downstream, with the demo moving from upload to extraction results and the report noting downloadable output with no extra transformation step."},{"id":"527704b6-b5a4-4c3d-ba5e-e7d4275464b9","tool":"retab","tool_name":"Retab","verdict":"worked","score":null,"score_total":null,"note":"Produces directly copyable JSON from the workflow, so the bank-statement extraction is immediately usable without a transformation layer after configuration."}]}