BACEN Judicial Automation
RPA for batch case closure with Fortra Automate
Batch closure of court cases on legacy systems with no API, every day at 7 am, with 95% less processing time.
IMPACT METRICS
Savings
2.5 FTE
ROI in 4 months
Efficiency
95%
less total time
Frequency
Daily
runs at 7 am
- Technologies
- Fortra Automate
THE PROJECT
RPA automation, built on Fortra Automate, of judicial case closures in the BACEN system for a financial institution. The robot validates each case and groups them into large batches to close them in bulk, with no manual case-by-case processing.
With image recognition to navigate legacy interfaces, a text and image record of every critical step and email alerts to the client's teams, the automation runs on business days from 7 am and has everything processed before the working day begins.
STRATEGIC IMPACT
- Batch processing
- Court case tracking
- Real-time notifications
- Instant scalability
HOW IT WORKS
Architecture
01Automation (RPA)
Secure login, scanning of pending cases, batch consolidation, bulk closures and result validation, with image recognition wherever the interface offers no selectors.
02Batch intelligence
Groups cases by type, value and status, works out the best batch size and splits into sub-batches when the system needs stability.
03Communication and alerts
Automatic emails to the teams the client defines, with the batch summary, the number of closures, run time, errors and final status.
04Governance and audit
A timestamped log of every action, screenshots before, during and after each closure, files organised by date and batch, and traceability for regulatory audits.
Step by step
- 1Scheduled start on business days at 7 am, with precondition checks.
- 2Secure login and session validation.
- 3Scan of pending cases and data extraction: case number, value and type.
- 4Batch consolidation based on business rules and system limits.
- 5Pre-validation of every case in the batch: pending items and blocks.
- 6Batch closures executed while monitoring the system's response.
- 7Case-by-case exception handling, without stopping the batch.
- 8Status check after each closure.
- 9Results consolidated into a run report.
- 10Email alert to the defined teams with a processing summary.
- 11Logs and evidence archived by date and batch.
KEY CHALLENGES
- Legacy systems (no API)How we solved it: Image recognition to locate buttons, fields and status messages dynamically.
- High volume of simultaneous casesHow we solved it: Consolidation into optimised batches, cutting processing time by more than 90% compared with handling cases one by one.
- Instability at peak hoursHow we solved it: Automatic retries with progressive back-off, and large batches split into sub-batches.
- Evidence for court proceedingsHow we solved it: Mandatory screenshots at four moments: before the run, during the closure, at confirmation and after processing.
- System-specific errorsHow we solved it: More than 15 error scenarios mapped, each with an automated recovery flow.
BUSINESS IMPACT
- 95% less processing time for closures, compared with manual case-by-case work.
- No more operational typing errors or wrongly selected cases.
- Processing guaranteed before the working day, freeing the team for higher-value work.
- Full judicial traceability, with documentary evidence for audits.
- Transparency for managers and control areas, with real-time alerts.
- Instant scale for volume peaks, without adding staff.
- ROI in 4 months, with 2.5 FTE redirected to exception analysis.
