Now with 142 AI Forensic Checks

Document fraud
stops here.

Upload any Indian document — Aadhaar, PAN, salary slips, certificates — and get an instant AI-powered authenticity verdict. Protect your company from forged documents.

0 AI Checks
<2min Per Document
0 % Detection Rate
Verification Result
SUSPICIOUS Tampering indicators detected
47/100
Mathematical Verification FAILED
Edit History Analysis FAILED
OCR Text Extraction PASSED
Font Consistency PASSED
Scanning...
AI Analyzing
Identity Match

Trusted by companies across India for document verification

Financial Services
HR & Recruitment
Banking & NBFC
Education
Real Estate
Legal Firms

Everything you need to
verify documents

From upload to verdict in under 2 minutes. Our AI engine runs 142 forensic checks across 11 verification categories.

142 AI Forensic Checks

Visual analysis, metadata forensics, image manipulation detection, content verification, and more — all automated.

PDF Edit Detection

Detects incremental saves, white-rectangle overlays, font mixing, creator/producer mismatches, and annotation tampering.

Salary Math Validation

Verifies Gross = Basic + HRA + DA, Net = Gross - Deductions, PF ~12% of Basic. Catches edited amounts instantly.

Auto Candidate Linking

Extracts name, DOB, PAN, Aadhaar from documents and auto-links to candidate profiles. Smart fuzzy matching.

4-Level Verdict System

Clear, Review Recommended, Suspicious, or Fraudulent — each with actionable next steps for your team.

Indian Document Support

Aadhaar, PAN, Passport, Driving License, Voter ID, Salary Slips, Bank Statements, Educational Certificates.

Three steps to
complete verification

01

Upload Document

Drop any PDF, PNG, or JPG. Our quality gate checks resolution, format, and readability before processing.

02

AI Analyzes

142 checks across visual, metadata, forensic, content, security, AI/ML, and database verification categories.

03

Get Verdict

Instant verdict — Clear, Review, Suspicious, or Fraudulent — with detailed findings and actionable recommendations.

Every trick they try,
we catch it

Our AI has been tested against real forged salary slips, edited Aadhaar cards, and fake certificates. Here's exactly what it detects — with live output from our engine.

Edited Salary Slip Uploaded

A candidate changed their salary from ₹25,000 to ₹48,323 using a PDF editor. They also altered the Net Pay and Total Deductions to look consistent.

Amount Changed White Overlay Font Mismatch
DocEnsure Result
SUSPICIOUS 50/100
MATH MISMATCH: Gross (48,323) ≠ Components (25,823)
4 PDF edit indicators: editor tool, %%EOF×3, white overlays
7 font families detected — payroll systems use 1-2
Verdict: Do NOT accept. Request original from employer.
Critical — New Threat

AI-Generated Document Detection

Candidates are now using ChatGPT, Midjourney, and Photoshop AI to create entirely fake documents from scratch. These AI-generated documents look perfect to the human eye — but our 6-signal neural forensic engine catches what you can't see: no scanner noise, uniform frequency spectrum, missing JPEG compression artifacts, and unnatural color channel correlations.

ai_detection.log
AI Generation Score: 23/100 — LIKELY AI GENERATED
├─ Frequency spectrum: suspiciously smooth (ratio 0.12, expected >0.4)
├─ Noise residual: too uniform (variation 0.3, real docs >1.5)
├─ Color channel correlation: 0.08 (cameras produce >0.3)
├─ Texture LBP entropy: 4.1 (real docs >6.5)
├─ No JPEG block boundaries — image was never captured by camera
└─ No scanner/device metadata found in PDF
Critical

Salary Math Validation

We verify the math. Basic + HRA + DA must equal Gross. Gross minus deductions must equal Net Pay. PF should be ~12% of Basic. If any amount was edited, the math breaks.

verification_output.log
MATH MISMATCH: Gross (48,323) ≠ Components
sum (25,823), difference: 22,500.00
Net pay mismatch: difference 1,300.00
Critical

Deep PDF Structure Analysis

Every edit leaves fingerprints. We inspect %%EOF markers, xref tables, white overlays, annotations, and editor tool signatures.

edit_history.log
4 indicators: editor tool, %%EOF×3,
white overlays, Creator/Producer mismatch
High

Font Forensics

Payroll systems use 1-2 fonts. When someone edits a value in a different app, the new font doesn't match. We extract every embedded font and flag mixing.

font_check.log
HIGH RISK: 7 font families in salary slip
ArialMT, Calibri, TimesNewRoman, Helvetica...
High

Image-Level Forgery

Our Error Level Analysis, noise pattern detection, and copy-move analysis expose pixel-level manipulation that humans can't see.

forensics.log
ELA variance 34.2 — different compression
Noise CV 2.68 — inconsistent across blocks
Smart

Cross-Document Identity

Aadhaar says "Rahul Sharma" but salary slip says "Rahul S"? Our fuzzy matching + DOB + PAN linkage catches identity gaps across all documents.

identity.log
MATCHED: "Rahul S" ↔ "Rahul Sharma" (87%)
DOB: exact match, PAN: confirmed on both
11Verification Categories
142Individual Checks
<2 minProcessing Time
4Verdict Levels

Simple, transparent pricing

Start free. Scale as you grow. No hidden fees.

Starter

Free/forever

For individuals and small teams getting started.

  • 25 documents/month
  • All 142 verification checks
  • PDF reports
  • Candidate profiles
Contact Us

Enterprise

Custom

For large organizations with compliance needs.

  • Unlimited documents
  • On-premise deployment
  • Custom integrations
  • SLA guarantee
  • Dedicated support
  • Audit & compliance
Contact Us

Verify any Indian document

From government IDs to financial records — our AI understands the structure, security features, and validation rules for every major Indian document type.

🪪

Aadhaar Card Verification

Validates QR code cryptographic signature, checks UIDAI format compliance, verifies Verhoeff checksum on 12-digit UID, and detects photo tampering on India's universal identity document.

💳

PAN Card Verification

Validates PAN format (ABCDE1234F), verifies entity type code, checks hologram region integrity, and cross-references name and DOB consistency across submitted documents.

💰

Salary Slip Fraud Detection

Validates complete payroll math — Basic + HRA + DA = Gross, Gross - Deductions = Net Pay, PF ~12% of Basic. Detects PDF editing tools, font mixing, and white-rectangle overlays used to change amounts.

🏦

Bank Statement Verification

Checks transaction pattern consistency using Benford's Law analysis, validates running balance calculations, detects round-number anomalies, and identifies PDF editor tool signatures.

🎓

Educational Certificate Verification

Verifies SSC, HSC, degree, and diploma certificates. Checks institutional patterns, board-specific formats, detects scanned document manipulation and AI-generated fake certificates.

📕

Passport & ID Verification

Supports Indian Passport, Driving License, and Voter ID. Validates MRZ codes, state RTO codes, checks expiry dates, security features, and detects image splicing on photo IDs.

Frequently asked questions

What documents can DocEnsure verify?

DocEnsure supports Aadhaar Card, PAN Card, Passport, Driving License, Voter ID, Salary Slips, Bank Statements, Educational Certificates (SSC, HSC, Degree, Diploma), Property Documents, and Legal Documents. Our AI adapts its 142 checks to each document type's specific structure, security features, and validation rules.

How does DocEnsure detect edited salary slips?

We use a multi-layered approach: Mathematical validation checks that Gross = Basic + HRA + DA, Net = Gross - Deductions, and PF ≈ 12% of Basic. PDF forensics detect editor tool signatures (iText, PDFtk, Foxit), incremental save markers (%%EOF), white-rectangle overlays, and content stream modifications. Font forensics flag when edited text uses different fonts than the original payroll-generated document.

Can DocEnsure detect AI-generated fake documents?

Yes. Our 6-signal neural forensic engine analyzes frequency spectrum (AI images lack natural sensor noise), noise residuals (AI produces uniform noise vs. varied real noise), color channel correlations (cameras produce correlated RGB noise), texture naturalness via LBP, JPEG compression artifacts, and PDF metadata. Documents from ChatGPT, Midjourney, DALL-E, and Photoshop AI are flagged instantly.

How long does verification take?

DocEnsure processes each document in under 2 minutes, running 142 forensic checks across 11 categories: Visual Analysis, Signature Verification, Metadata Forensics, Image Forensics, Content Verification, Security Features, Database Verification, AI/ML Detection, Audit & Compliance, Document Liveness, and Biometric Verification.

What verdict levels does DocEnsure provide?

Every document receives one of four actionable verdicts: Clear (authentic, no action needed), Review Recommended (minor concerns, manual check advised), Suspicious (significant tampering indicators, don't accept without investigation), or Fraudulent (strong evidence of forgery, reject immediately). Each verdict includes detailed findings and specific next-step recommendations.

Which industries use DocEnsure?

DocEnsure is used by HR & recruitment firms for background verification, banks & NBFCs for KYC compliance, financial services for loan application verification, educational institutions for admission document checks, real estate companies for property document validation, and legal firms for evidence authentication.

How is candidate identity linked across documents?

DocEnsure automatically extracts name, DOB, PAN, Aadhaar number, and employee ID from each uploaded document using OCR and regex extraction. It then matches against existing candidate profiles using exact ID matching (PAN/Aadhaar), fuzzy name similarity (handles "Rahul S" vs "Rahul Sharma"), DOB correlation, and even filename analysis. Documents are automatically linked to the correct candidate profile.

Stop accepting forged documents.

Get in touch to protect your hiring process with AI-powered document verification.

Get in Touch