Peta ePenyata
Per-bank parsers
Dedicated layouts for Maybank Islamic, CIMB Savings, Public Bank, Hong Leong, AmBank and RHB. Handles password-locked PDFs, multi-page screenshots from mobile apps, and OCR of scanned printouts.
Built in Malaysia · Pilot open
Read statements. Verify payslips. Done.
Malaysian document intelligence for lenders, accountants, HR and SMEs. Drop a bank statement — we extract every transaction, reconcile every running balance and surface recurring commitments. Drop a payslip — we verify gross, deductions (EPF / SOCSO / EIS / PCB) and net take-home. Every result is confidence-tiered: verified, or flagged for review — never faked.
Statements + payslips
Lenders, accountants, HR, SME
Row-level
Confidence and flags
~30s
Per statement
PDF / XLS / CSV / JSON
Dashboard and API
Live today for Malaysia's Tier 1 banks. Tier 2 and Islamic windows next on the roadmap.
Maybank
CIMB
Public Bank
Hong Leong
AmBank
RHB
Bank Islam (soon)
Bank Rakyat (soon)
Built for the way Malaysian banks actually issue statements.
We tuned a parser per bank, against real statements your customers download from Maybank2u, CIMB Clicks, HLB Connect and the rest. Nothing here is a generic LLM wrapper.
Reconciled balances
Every row carries a confidence score and validation reason. Running balances are checked end-to-end so a misread cent never reaches your underwriter.
Lender snapshot
Average monthly income, income consistency, cashflow volatility, low-balance days, and an affordability band with the exact reason the engine landed on it.
Recurring commitments
Pulls out rent, TNB, Astro, PTPTN, loan instalments, insurance, takaful, and SpayLater. Debt-service ratio computed against recognised income, not gross inflows.
Credit-committee PDF
One branded page your credit team can attach to a file. CSV and Excel exports include the lender insights, monthly series, and recurring commitments sheets.
JSON API for fintechs
Same pipeline as the dashboard, callable from your origination or KYC flow. Transactions, validation flags, insights and lender snapshot - every field with explainable confidence.
Three steps from PDF to underwriting decision.
-
1
Upload
Drop the bank PDF or upload screenshot pages. Single statement or multi-month bundle.
-
2
Extract & reconcile
Bank detected, transactions parsed, balances reconciled, and confidence scored row by row.
-
3
Report
Branded PDF for credit committee, Excel for analysts, CSV for the data team, JSON for your stack.
Pricing built for Malaysian lenders.
Start with a paid pilot. Scale by statement volume or API call. No surprise overages.
Starter
RM 299/month
SME finance teams
- Up to 100 statements / month
- CSV & Excel exports
- Dashboard for 2 users
- Email support
Business
RM 1,499/month
Underwriting teams
- Up to 1,000 statements / month
- Branded lender PDF report
- Bulk upload & recurring insights
- Team dashboard, audit logs
- Priority response
Enterprise / API
RM 5,000+/month
Fintech & SaaS
- API with SLA
- Custom adapters & categories
- Private deployment available
- PDPA & processor terms
Built with security and PDPA in mind.
Account masking
Account numbers are masked in exports and the lender report by default.
No training on customer data
Statements analyzed in your workspace are never used to train shared models.
Workspace isolation
Each workspace has its own session and storage. Audit logs and retention are roadmapped.
On-prem option
Enterprise customers can run the analyzer inside their own VPC, no statement data leaves.
See your first statement parsed in 60 seconds.
Spin up a workspace and drop your first statement. Reconciled transactions, recurring commitments and a lender PDF land in your dashboard while you grab a coffee.