FOR LENDING & NBFCs

Process loan files at disbursal speed

OCR bank statements, extract income and KYC data to clean JSON, and redact PII before it touches your underwriting stack. Pay per page, scale with your book.

Get API key Read the docs

Async + webhook for high-volume batches Documents auto-purged after processing

LOAN-FILE PROCESSING

Turn a borrower folder into structured data

A single loan file is a stack of PDFs, scans and photos. These APIs read them, pull the fields your credit model needs, and strip identifiers your team should not store.

Ingest the file

Submit PDF, TIFF, JPG, PNG or scanned multi-page images. Pre-processing handles deskew, denoise, orientation and contrast on poor-quality scans.

OCR bank statements

OCR Full-Text reads statements, salary slips and ITRs into searchable text and a searchable PDF, including Hindi, Tamil, Telugu and other Indian scripts.

Extract income & KYC

Extract Pro returns named fields, confidence scores and bounding boxes from bank statements, KYC forms and payslips as JSON your underwriting flow can consume directly.

Redact PII

PII Masking redacts Aadhaar, PAN, account numbers, names and addresses while preserving the original layout, so the stored copy is audit-safe.

Built for NBFCs, MFIs and fintech lenders

Whether you disburse a hundred loans a month or a hundred thousand, the work is the same: read the documents, get reliable fields, keep PII controlled. The cost moves with your disbursal volume, not a fixed seat licence.

Flat per-page pricing with no surge, tiered overages or per-call ceilings.
Extract Pro adds confidence scores, guardrails and validation rules for regulated credit workflows.
Synchronous mode for quick eligibility checks; async with webhook callback for overnight batch runs.
p95 latency under 8s for a 10-page document, backed by a 99.5% uptime SLA.
Hosted in India (GCP Mumbai) for Indian data; Johannesburg for Africa and the Middle East.

Explore Extract Pro →

RELEVANT APIS

The three APIs most lenders start with

Mix and match per page. Each call returns structured JSON with extracted data, confidence and bounding boxes.

OCR Full-Text

Full text extraction plus searchable PDF output from statements, slips and scanned KYC. Multi-language including Hindi, Tamil, Telugu and Arabic. Rs 0.25 /page (about $0.003 / page).

OCR Full-Text →

Extract Pro

Structured fields from bank statements, KYC forms, payslips and invoices, plus confidence scores, guardrails and validation rules for regulated workflows. Rs 1.40 /page (about $0.017 / page).

Extract Pro →

PII Masking

Redact Aadhaar, PAN, account numbers, names, addresses, phone and signatures with configurable mask rules. Output preserves the original layout. Rs 0.25 /page (about $0.003 / page).

PII Masking →

Scaling spend matches scaling volume

Start on the one-time Rs 500 free trial credit (no card) for roughly 1,000+ OCR or masking pages, around 550 Extract Basic, or 350 Extract Pro pages. As your book grows, move to top-up packs from Mini Rs 10,000 up to Scale, where larger packs add bonus credit. Credit does not expire while your account is active, so seasonal disbursal peaks never strand prepaid balance. See full rates and pack bonuses on pricing.