DynamicForm Start free trial
Step 11 of 15 · Document Analyzer · topic 2 of 3

How it works

How does the Document Analyzer work?

Identity documents

Aadhaar (front and back joined, number masked), PAN and Voter ID are read into the form.

Face match

The photo on the ID is compared with a live selfie, giving a match percentage.

Bank statements

Text PDFs are parsed in milliseconds; scanned pages are read by a vision AI model. Every transaction keeps its page number and is categorised.

Fraud checks

Edited-PDF detection, round-tripping of money, sudden large deposits, money parked and withdrawn, balance continuity, bounces.

Business documents

GST returns, balance sheets and profit-and-loss statements mapped to a standard structure.

The ID photo is matched with a live selfieAn identity card photo and a live selfie on a phone, joined by a ring showing a 94 percent match.XXXX XXXX 4821ID card photo94%matchLive selfiePhoto on the ID compared with the selfie
ID photo and selfie side by side, joined by a match ring filling up.
Twelve months of money in and out, at a glanceMoney coming in drawn as a teal band above the line and money going out as a coral band below it, with the running balance as a blue line; example values.InOutBalance keeps risingMayJunJulAugSepOctNovDecJanFebMarApr12 months read from 2 statements · example values, ₹ lakh
Monthly inflow and outflow as a flowing river with the balance level.
Example · Customer Loan Origination

In the loan example, ID and selfie uploads fill the onboarding form, bank statements fill the financials tab, and any red flag sends the case to manual review.

See the full loan example

Build your first app free

Up to 10 users. 30 days to build, then 30 days live after go-live.

Start free trialTalk to sales

Updated 30 September 2026