Step 1 — Upload the Bank Statement PDF
Drag and drop or click to upload any bank statement PDF — from Chase, Wells Fargo, Bank of America, credit unions, or any US bank. Supports files up to 20MB and statements covering up to 24 months.
- Accepts any PDF layout — no special format needed
- Chase, Wells Fargo, BofA, Citibank, credit unions, regional banks
- Up to 20MB, up to 24 months of history
- Drag-and-drop or file picker — works on desktop and mobile
Step 2 — AI Reads Every Transaction
Our model reads every deposit, withdrawal, fee, and transfer on every page. It understands the context — it knows the difference between a payroll deposit, an ACH transfer, an MCA repayment, and a refund. It doesn't skip lines.
- Every transaction extracted — no skipping, no sampling
- Identifies transaction type, merchant, amount, and date
- Distinguishes income from transfers from refunds
- Works on digital PDFs (99%+ accuracy) and scanned PDFs
Step 3 — Deep Lender Analysis
Once transactions are parsed, StatementScrub runs the same checks an experienced underwriter runs manually — but in seconds.
- Average monthly income calculated and verified
- Income sources identified by name and type
- NSF events counted month by month
- MCA repayments flagged with merchant name + estimated daily payment
- Fraud patterns detected (structuring, cycling, velocity spikes)
- Risk score 0–100 with APPROVE / REVIEW / DECLINE
Step 4 — Receive Your Lender Report
You get a structured report ready for your loan file. It contains everything your underwriting decision requires — no manual cross-referencing, no second-guessing.
- Risk score 0–100 with Low / Medium / High / Critical band
- APPROVE / REVIEW / DECLINE with AI plain-English reasoning
- Month-by-month cash flow table
- Income verification with source names
- MCA detection with merchant and daily payment estimate
- Export as PDF (print) or CSV for your spreadsheet
Full Detection Suite
Every Check in Every Report
12 automated checks — the same things a trained underwriter looks for, done in seconds on every statement.
Sample Report Output
What a Report Actually Looks Like
This is real output structure — not a mock-up. Every field below is populated from the bank statement you upload.
Bonus Feature — Any Device
Analyze Statements via Telegram
Send a PDF to @StatementScrubBot on Telegram and get the full report back in chat — income, NSF count, MCA detection, risk score, and recommendation. Works on any device, no browser needed.
Get Access →Income: $9,840/mo
NSF: 0 · MCA: None
Risk: LOW (Score 18)
→ APPROVE
See It With Your Own Statements
Start with 3 free reports — no credit card, no setup, no commitment.