рд╕рднреА рдЯреЗрдореНрдкрд▓реЗрдЯ

Bank Reconciliation

RobomotionтАв4 рдорд╛рд╣ рдкрд╣рд▓реЗ рдЕрдкрдбреЗрдЯ рдХрд┐рдпрд╛ рдЧрдпрд╛
Bank Reconciliation

рдУрд╡рд░рд╡реНрдпреВ

Takes the payments and receipts the ERP could not reconcile and confirms each one against the real bank statement. It pulls the ERP's flagged paid-bill exceptions and open invoices awaiting cash, then signs in to the bank, searches the operating account by reference and reads what actually settled - quantifying the four ways money and ledger drift apart: a missing payment, a transposed amount, a double payment and unapplied cash. Writes a reconciliation CSV. The story is evidence: the ledger raises the question, the bank statement answers it, and every discrepancy comes out with a number attached.

Bank Reconciliation

The ledger says a bill was paid. The bank is where you find out whether it actually was тАФ at the right amount, once, and only when it should have been. Reconciliation is the control that catches the gap between the two, and it is exactly the kind of patient, evidence-gathering work a robot does better than a person at month-end: for every item in question, pull the statement, read what settled, and put a number on the discrepancy.

This flow does that. It takes the payments and receipts the ERP could not tie out and confirms each against the operating account, classifying the four ways money and ledger drift apart.

It spans two fictional training systems тАФ RAP One, the ERP, and AcmeBank, the bank тАФ whose data is entirely synthetic and reconciles to the cent.

What it produces

bank-reconciliation.csv in your home folder: every flagged item with its verdict and the exact figures from both sides. On the seeded data that is five confirmed discrepancies:

  • Missing payment тАФ NW-2026-0442 is Paid in the ledger but no debit left the bank; тВм4,998 outstanding.
  • Transposition тАФ HE-2026-06-4471 settled at тВм1,824.33 against a ledger amount of тВм1,842.33 (a transposed digit, тВм18.00 out).
  • Double payment тАФ VF-2026-06-8801 was paid twice: two debits of тВм489.90.
  • Unapplied cash тАФ INV-2026-0641 (тВм8,901.20) and INV-2026-0655 (тВм6,330.80) were received in the bank but the invoices are still open in the ledger.

How it works

1. Pull the ledger's open questions

Sign in to the ERP. Its exceptions worklist already flags the paid bills it could not tie out, and the customer invoices where cash is expected but the item is still open. The robot reads the reference and amount for each тАФ the items to prove against the bank.

2. Prove each against the bank

Two-factor sign in to the bank and open the operating account. For every flagged item, search the statement by its reference and read what actually settled тАФ no debit, a debit at the wrong amount, two debits, or a credit that arrived. The bank statement is the independent record; the count and amount decide the verdict.

3. Classify & report

Zero debits is a missing payment; one debit at the wrong amount is a transposition; two debits is a double payment; a credit against a still-open invoice is unapplied cash. Each verdict carries the figures from both sides, written to a CSV a finance team can action.

Running it

Ready to run as-is. It signs in with the published training credentials (HTANAKA for the ERP, hiroshi.tanaka@globex.example with a static one-time code for the bank), which are not secret. Real system credentials belong in the Robomotion Vault, never in a flow.

Web Automation рдХреЗ рдФрд░ рдЯреЗрдореНрдкрд▓реЗрдЯ

рд╕рднреА 20 рджреЗрдЦреЗрдВ тЖТ