CASE STUDY · TOLARAM GROUP · AUTOMATION
A Month of Matching in Minutes: Bank Reconciliation Bot
Finance spent the whole month matching bank entries by hand. The bot I built does it in 5–10 minutes.
CONTEXT
At the Tolaram group, finance reconciled bank statements against the ledger in Oracle E-Business Suite every month. My team lead initiated the project; I designed and built the bot.
THE CHALLENGE
Matching entries by hand took the finance team the entire month.
WHAT I BUILT
- Download the statementOnce a month, the Automation Anywhere bot logs in to the bank website and downloads the statement. Bank logins are kept in Automation Anywhere's credential vault.
- Pull the ledgerIt connects to the database and pulls the ledger from Oracle E-Business Suite.
- MatchIt matches entries on transaction amount and description.
- Report the exceptionsEvery mismatch goes into an Excel file, emailed to the responsible finance team for review.
RELIABILITY
- Credentials stored in the credential vault, never in the bot's script.
- The bot never forces a match: anything uncertain goes to a person.
- Clear output: finance gets only the exceptions, not the full ledger.
ARCHITECTURE · AS BUILT
OUTCOME
Reconciliation that took finance the entire month now runs in 5–10 minutes, and the team only reviews the exceptions.
SKILLS SHOWN
Certified: Automation Anywhere Advanced RPA Professional (V11.0), 2020
← Back to all projectsCONTACT
Let's talk.
Work, photography collaborations, or just hello.