Financial Document
Screening & Verification
Automatically detect internal financial leakage and manipulated financial profiles. Upload bank statements, payslips, or financial reports and surface document anomalies in minutes.
Built for Lenders, Internal Auditors, and Compliance Teams
Three modules.
One dossier.
FinGuard runs screening, reconciliation, and forensics against a single case file (dossier), so every review is complete.
Screening
Extract structured data from every financial document and evaluate it automatically against ~70 deterministic and probabilistic rules.
- Supports bank statements, payslips, tax filings, KTP/NPWP, and financial statements
- Arithmetic, structural, content, behavioural, and forensic layers per document
- Per-field confidence, evidence quotes, and links back to the source document
Reconciliation
Spot mismatches between documents instantly. Reconcile identities, balances, and transactions across the whole file.
- 14 cross-document rules for names, dates, balances, amounts, and recurring income
- Tolerances by field type: exact, fuzzy match, or numeric variance
- Side-by-side evidence with highlighted diffs and one-click dismissal
Forensics
Deep tamper detection on every uploaded PDF, so manipulated documents are caught before they cause losses.
- ~30 forensic rules calibrated against a corpus of genuine bank-issued PDFs
- Detects font substitution, editor-application traces, and embedded XMP metadata
- Watermark integrity against genuine patterns, plus perceptual image-hash collisions
A transparent engine,
step by step.
FinGuard is not a black box. Every flag records the exact rule version that produced it.
Ingest
Upload financial documents (KTP, NPWP, payslips, bank statements, financial statements) through the secure API or the FinGuard dashboard.
Classify
The Classifier worker identifies each document type and checks file-format integrity (PDF, image, and more).
Extract
The Extractor worker pulls out text, table structure, and key fields using purpose-calibrated OCR models and AI processing.
Evaluate
The Evaluator worker checks the extracted data against 150+ anomaly rules across 5 detection layers.
Profile
The Profiler worker combines every finding into a final risk profile with a complete, ready-to-review audit trail.
Built for verification teams, not transactions.
Screening analysts, audit, compliance, internal investigations, and external examiners all work from the same evidence in one integrated case file.
- Quarterly file reviews at scale (sampling and audit)
- A complete, immutable audit trail
- Thorough screening of vendors, new business partners, and customers
- Deep single-file forensics to detect engineered PDFs
Audit-ready trail
Auditors and admins can query audit logs straight from the UI. Logs are kept for 5 years (1 year online, 4 years in cold storage) to meet audit requirements.
Security and UU PDP compliance
Hosted on Biznet Cloud in Jakarta. Data is envelope-encrypted with KMS, and entity identities are automatically pseudonymised before anything reaches an external AI model.
Pay only for
the pages you analyse.
One page is one analysed document page. No seat fees and no platform minimum. All three modules are included.
Starter
- Prepaid packs of 100, 300, or 500 pages
- Full screening, reconciliation, and forensics modules
- Up to 3 users
- 30-day file retention
- Community support
Growth
- Prepaid packs of 1,000, 2,500, or 5,000 pages
- Bundle savings vs pay-as-you-go
- Up to 25 users
- 5 GB shared file storage
- Custom retention up to 90 days
- Email support with 8-hour SLA
Enterprise
- Prepaid packs of 10,000, 20,000, or 50,000 pages
- Dedicated tenant
- Custom retention up to 5 years
- Custom-branded reports, API, and webhooks
- Dedicated CSM and 99.9% SLA
- On-premises installation (phase 3)
Your first 10 pages are free, with no commitment. Upload a file and see what FinGuard finds.
Ready to secure your transactions?
Start your free trial with 10 pages of financial documents in the FinGuard workspace today.
Let's Build the Future Together
Ready to transform your organization with AI? Get in touch.