01 · HANDWRITINGlive demo
Handwritten cheques, not just printed text.
Cursive payees, rushed signatures, stylised dates. Our field-specific models are optimised for and evaluated on real cheques.
Per fieldconfidence and review state
cursive readySubmit front and back check images and receive structured JSON for MICR, amounts, payees, dates, signatures, endorsements, confidence scores, and validation states. Chequedb gives developers synchronous responses, webhook-driven review events, SDK options, and cloud or on-premise deployment paths.
This is the developer integration overview for teams evaluating a cheque OCR API, check reader OCR response, or bank check parser. Continue to the API documentation for endpoint-level request formats, or review cheque data extraction for broad field coverage and operational use cases.
Developer proof
Chequedb returns field values with confidence, validation results, review reasons, and workflow status instead of raw OCR text. Developers can post accepted items to an ERP, hold low-confidence fields for review, or receive webhook events when an exception is resolved.
{
"check_id": "chk_01j2r9m4k8",
"status": "review_required",
"fields": {
"micr": {
"routing_number": "021000021",
"account_number": "1234567890",
"check_number": "1042",
"confidence": 0.99
},
"amount": {
"courtesy": "1850.25",
"legal": "one thousand eight hundred fifty and 25/100",
"match": true,
"confidence": 0.96
},
"payee": { "value": "Summit Supply LLC", "confidence": 0.87 },
"date": { "value": "2026-03-20", "status": "valid" }
},
"review": {
"required": true,
"reasons": ["payee_confidence_medium"]
},
"webhooks": [
"check.review_required",
"check.review_resolved"
]
}Three capabilities almost nobody else ships together — handwriting trained on real cheques, a parser that understands how people actually say money, and deployment that keeps every account number inside your network.
Cursive payees, rushed signatures, stylised dates. Our field-specific models are optimised for and evaluated on real cheques.
“Lakh”, “naira only”, “k”, “no cents”, “no paisa”. Our parser handles regional financial language and reconciles it against the numeric box every time.
Cheque images carry account numbers, signatures, and payment history. Deploy Chequedb inside your infrastructure so image and field data can remain within the configured environment.
Generic cloud OCR can read text. Almost nobody combines cheque-specific handwriting recognition, colloquial amount reconciliation, and on-prem deployment for non-bank businesses. That’s the gap ChequeDB fills.
Book demoThe integration accepts front and back check images, identifies the cheque fields, and returns a structured response rather than a block of OCR text. MICR supplies routing, account, and check-number control data; OCR and ICR read visible printed and handwritten fields.
Each JSON field can include its value, confidence, validation result, and review state. That lets an application accept a trusted value, hold an uncertain field, or present the relevant image region to a reviewer without inventing its own response contract.
Synchronous clients can act on the immediate response. Asynchronous workflows can consume webhook events when an item needs review or a decision is resolved. For a field-by-field explanation before choosing an integration, see bank check OCR and cheque data extraction.
Return the fields and validation signals downstream applications need for review, posting, and exception handling.
| Field Group | Data Returned | Why It Matters |
|---|---|---|
| MICR and bank data | Routing number, account number, check number, bank identifier | Supports posting, routing validation, duplicate checks, and downstream workflow matching. |
| Amount and payee | Courtesy amount, legal amount, payee name, memo region | Helps teams compare numeric and written values, match payee rules, and catch mismatch-driven exceptions. |
| Date and image quality | Issue date, stale or future-date flags, blur and framing checks | Gives the client explicit signals for holding unusable or policy-sensitive images before a deposit workflow continues. |
| Signature and endorsement | Signature region, endorsement presence, review-ready image coordinates | Gives fraud, operations, and reviewer tools image context for manual checks and approval routing. |
A predictable integration path from image submission to a structured response and downstream workflow event.
Submit front and back images from a mobile app, scanner-based RDC flow, teller desk, kiosk, or back-office client.
Read MICR and visual fields, localise cheque signatures and endorsements, and return confidence-aware JSON.
Score image quality, compare amounts, validate routing, flag duplicates, and route exceptions to reviewers.
Return the current state in JSON and emit webhook events when review or deposit states change. Continue to the API documentation for endpoint details.
{
"check_id": "chk_01hy9v3x8c",
"status": "review_required",
"fields": {
"routing_number": { "value": "021000021", "confidence": 0.99 },
"account_number": { "value": "1234567890", "confidence": 0.98 },
"check_number": { "value": "1042", "confidence": 0.99 },
"amount": { "value": 1850.25, "confidence": 0.96 },
"payee": { "value": "Summit Supply LLC", "confidence": 0.87 },
"date": { "value": "2026-03-20", "valid": true }
},
"checks": {
"image_quality": "pass",
"duplicate_risk": "low",
"amount_match": "pass",
"endorsement_present": true
}
}Start with the delivery model that matches your application surface and control requirements.
Built for operations teams that need field-level confidence signals, exception traceability, and flexible deployment.
The same structured fields that feed deposit and review workflows can push directly into SAP, Odoo, QuickBooks, Xero, or custom ERP systems via the Chequedb REST API — with idempotent posting, control totals, error handling, and full audit trail synchronisation. The extracted MICR account number determines the GL account; the cross-validated amount posts as the transaction value; the payee name and date populate the journal entry reference.
The same OCR layer across mobile, scanner, and back-office check capture.
Use the API for integration, then connect the same extracted fields to review, archive, and management workflows.
Broad field coverage for printed and handwritten cheque OCR, validation, and operational review.
Route low-confidence OCR results into approval and exception queues.
Capture quality-controlled cheque images from scanners and mobile flows.
Store extracted fields, images, statuses, and audit history for search.
Book a technical walkthrough to review field-level JSON, confidence handling, webhook states, SDK options, and deployment fit. Use the API documentation for endpoint-level request patterns.