WSA Billing System Enhancement: BAUT vs Billing Data Comparison
Overview
The enhancement introduces a robust data-comparison engine that automatically aligns BAUT (Berita Acara Uji Terima) transaction logs with WSA Billing records. By ingesting Cycle-1 and Cycle-2 billing tables, the system surfaces mismatches in real time, classifies anomalies by root cause (e.g., WorkCode gaps, quadrant mis-flags, CT-0 states), and drives rapid corrective actions across Telkom, Telkomsel, DIT, and TSSC teams.
Key Features
Automated Record Matching: Utilises fuzzy-key joins on Order ID, Service ID, and STO codes to pair BAUT events with corresponding billing rows, eliminating manual VLOOKUP reconciliation.
Discrepancy Taxonomy & Scoring: Applies rule-based logic to tag variances (Missing, Duplicate, Mis-categorised, Amount Delta) and assign financial risk scores for prioritised resolution.
Interactive Grafana Dashboard: Real-time widgets display variance heat-maps, trendlines by product segment, and drill-downs to the offending WorkCode or Billing Status.
Root-Cause Drill Workflow: One-click export triggers back-end JIRA tickets that pre-fill anomaly context, expediting hand-offs to DIT developers or TSSC DBA teams.
Technologies Used
Oracle 19c & Geneva: Primary data stores for Cycle-1 and Cycle-2 WSA Billing tables, accessed via materialised views.
Grafana + PostgreSQL: Serves aggregated discrepancy metrics with role-based access control for finance, QA, and engineering stakeholders.
ServiceNow/JIRA API: Automates incident creation from dashboard actions, embedding anomaly metadata for quicker triage.
Challenges and Learnings
Heterogeneous Schemas: BAUT XML extracts and Billing relational tables required schema harmonisation; JSON flattening and dynamic SQL views resolved structural drifts.
Latency vs Accuracy Trade-off: Initial hourly refresh caused outdated variances; implementing partial CDC with Oracle LogMiner reduced end-to-end delay to <5 minutes.
High Cardinality Join Keys: Composite indexes on Order ID + Billing Month cut reconciliation runtime from 42 min to 7 min for 18 million monthly rows.
Change Management: Cross-unit adoption demanded extensive stakeholder workshops and SOP updates, highlighting the value of visual dashboards in driving cultural shift.
Outcome
Since rollout, the solution detects billing-BAUT mismatches within T + 5 minutes, reducing revenue-at-risk by an estimated ₨ 4.8 billion per month and slashing manual reconciliation effort by 80 %. The modular architecture supports upcoming IT system enhancements (IBSS logging fixes, new Geneva tables) without re-engineering, positioning Telkom's WSA Billing for sustained accuracy and audit-readiness.
