Drop a CSV or a text-based PDF statement, or load the labelled sample. Merchants are matched against a public brand list, and unusual debits are flagged. Parsing happens in your browser. This demo never talks to a bank, and it will not read filenames from a private statement folder.
The bundled sample is invented. Expected flags: a ₹48,500 UPI to an unknown merchant, a ₹2,950 food order about 6.7× the other food spends, and two identical Flipkart debits a day apart. A large BigBasket shop also flags against smaller grocery trips.
Transactions
—
Flagged
—
Closing
—
By category
| Category | Txns | Debits (₹) | Credits (₹) |
|---|
Transactions
| Date | Merchant / narration | Category | Debit | Credit | Balance | Flags |
|---|