Business documents
Capture text from invoices, delivery challans, textile roll sheets and forms. Preserve raw values when a date, quantity or unit is ambiguous.
Documents in. Useful data out.OCR API: your documents,
OCR API: your documents,
ready for your software.
Turn scanned pages into text and structured JSON. Review every field against its source, then connect the data to the systems your business already uses.
Private workspace · Prepaid credits · Source-linked results
invoice.pdf → structured data
INVOICEILLUSTRATIVE DOCUMENT · NOT CUSTOMER DATAValues linked to original page coordinates
Invoice no.INV-0241
Date08 Oct 2026
DESCRIPTION QTY AMOUNTCotton fabric120 m
Lot referenceLT-084
Total₹ 24,600.00
{
"invoice_no":
"INV-0241",
"total": {
"raw_value":
"24,600.00",
"page": 1,
"review_status":
"review_required"
}
}One asynchronous API. The documents you work with.
PDFJPEGPNGTIFF
Built for the handoff
From paperwork to a clear next step.
OCR reads the characters. Structured extraction organizes them. Source evidence helps your team decide what to trust and what to review.
Review with context
Inspect the original page beside extracted text, fields and tables. Select a value to see its source, record corrections and keep the automatic result.
A developer workflow
Submit once, poll the job, retrieve JSON. Scoped API keys, idempotent uploads and usage records make integration practical.
A small API. A complete workflow.
Upload. Extract. Review. Integrate.
Send your document
Upload an image or scanned PDF with a processing mode, language hints and an optional extraction schema.
Let the worker do the work
A durable background job validates and processes each page. Poll real page progress without blocking your application.
Keep the evidence
Retrieve text and structured data with page polygons, uncertainty and review status. Export JSON, UTF-8 text or table CSV.
Clear about capability
Readable evidence comes first.
English and Tamil printed-text adapters are included. Quality varies with the scan, typeface and page layout. Separate language benchmarks are required before publishing accuracy claims.
Handwriting, mixed handwriting and difficult photos require independent validation. The initial service does not claim verified handwriting recognition or signature verification.
Read the capability notesPrinted English & Tamil
Benchmark pendingConfigured OCR engines; verify against your own document samples.
Scans & phone photographs
Review advisedQuality checks flag blur, faint text and other detected limitations.
Tables & structured fields
Conservative extractionAvailable evidence is linked to pixels; complex or merged tables need review.
Handwriting & signatures
Not verifiedNo universal handwriting support. Never infer names from signatures.
Made to connect
One upload starts the job.
Keep your API key on your server. Use the browser playground with your workspace session, then bring the same processing flow into your application.
curl -X POST https://api.yulogo.com/v1/documents \
-H "Authorization: Bearer $YULOGO_OCR_KEY" \
-H "Idempotency-Key: $(uuidgen)" \
-F "file=@invoice.pdf" \
-F "mode=balanced" \
-F "languages=eng,tam"
# HTTP 202: poll the returned job URL.
# The API hostname requires deployment.OCR API questions
What is the difference between OCR and document intelligence?
OCR recognizes characters on a page. Document intelligence organizes text into fields or tables, links the result to source evidence and checks it against your schema. Business validation and human review are separate steps before posting data to an ERP or finance system.
Can I use the API for invoice processing?
Yes. You can submit invoice PDFs or scans and request text or structured fields. Test invoice numbers, suppliers, dates, totals, taxes and line items on representative documents. Ambiguous values should remain in a review queue rather than being guessed.
Is the invoice shown on this page a customer result?
No. It is an illustration with sample values, showing the document-to-JSON workflow. No benchmark accuracy or customer outcome is claimed. Evaluate your own samples in an authenticated workspace before rollout.
How does integration with existing software work?
Your server submits a document with a scoped API key, polls the asynchronous job and retrieves the result. Validate reviewed fields before writing them to your ERP, CRM or accounting application. Keep API keys on the server.
Can YULOGO help with a custom document workflow?
Yes. Share the document types, required fields, monthly volume and destination system in an API enquiry. We can discuss extraction, validation, review and integration requirements. Use approved or redacted samples when assessing sensitive documents.
Your documents, your workspacePrivate by design.
Private by design.
Reviewable by default.
Documents, keys, jobs and usage belong to your workspace. Downloads require authentication or short-lived storage access. Retention is configurable, and originals can be deleted from your history.
External OCR providers are disabled by default. Customer documents are not used for training without consent. Deployment and backup retention are documented separately.
Put your next document to work.
Start in the playground. Build with the API when you’re ready.