Key points in 30 seconds
- Receipt adjustments were keyed in by hand in Oracle EBS, with a slow login chain (browser, MFA, desktop app).
- An RPA robot automates login, SQL extraction, validation, reconciliation, remittance and reporting.
- Result: 100% field-level accuracy and 3 automated adjustment paths.
- Statuses are updated in real time via API, with a PDF and email report after every run.
Results in figures
100%field-level accuracy
3automated adjustment paths
Real timestatus tracking via API
Modularreusable sub-workflows
From login to final report
- Initial cleanupLeftover Oracle and Excel sessions closed, temp folders cleared.PowerShell
- Secure loginAutomated login, including MFA and the "AR_Manager" responsibility.RPA · Oracle EBS
- Data extractionDeposits pulled directly from the SQL database.SQL
- Date validationOut-of-range maturity dates filtered out and logged as failures.Pre-processing check
- ReconciliationAmounts compared, zero, positive or negative adjustment applied.Oracle EBS
- Remittance and reportReceipts batched, statuses updated via API, PDF and email report.API · Email
Entry and reconciliation: before / after
| Issue | Before | With the robot |
|---|---|---|
| Manual entry | Adjustments typed by hand in Oracle EBS: typos and financial discrepancies. | Data taken from the SQL database, with 100% field-level accuracy. |
| Authentication | Browser, mobile MFA and Oracle app: a slow, frustrating chain. | The whole login chain automated, MFA included. |
| Reconciliation logic | Mental calculation for each case: match, positive or negative adjustment? | Built-in rules: the right path (zero, positive, negative) applied instantly. |
| Receipt matching | Hundreds of Oracle receipts to match with deposits: major delays. | Bulk matching and batch creation in seconds. |
Reliability and control: before / after
| Issue | Before | With the robot |
|---|---|---|
| Maturity dates | Out-of-range dates processed, then reversed by hand. | Upfront validation: invalid dates filtered out and logged. |
| System stability | Leftover Oracle or Excel processes: hangs and crashes. | "Force Cleanup" at start-up, for a stable environment. |
| Visibility | No real-time tracking before the manual end-of-week report. | Statuses updated via API and a PDF/email report after every run. |
| Scalability | Logic rebuilt for each new scope. | Reusable modular sub-workflows, from one entity or region to the next. |
Reliable, tracked reconciliation
- No more entry errors: data flows from SQL to Oracle EBS with no re-keying.
- Instant reconciliation: hundreds of receipts matched in seconds.
- A stable environment: automatic cleanup before every run.
- Full traceability: every failure is logged with its reason, backed by a report.
Frequently asked questions
Can an RPA robot log into Oracle EBS with MFA?
Yes. In this project, the robot automates the whole login chain, including multi-factor authentication and selecting the "AR_Manager" responsibility.
How does the robot choose the adjustment type?
It compares the deposit amount with the receipts, then applies the matching rule: no adjustment, a positive adjustment or a negative adjustment.
How are failed deposits tracked?
Each failure is logged with its reason. Statuses are updated via API in real time, and a PDF report is emailed after every run.