LEVIATHAN is an enterprise-grade Core Banking Settlement and Reconciliation Engine designed to showcase the absolute precision and reliability of COBOL in modern financial infrastructures.
While 80% of global daily transactions still rely on COBOL, modern developers often overlook its unique ability to handle high-volume batch processing with zero margin for error. LEVIATHAN bridges this gap by integrating a high-precision COBOL core with a modern Cloud-native Python gateway.
In modern languages like Python or JavaScript, floating-point arithmetic (0.1 + 0.2) can lead to subtle decimal errors. In banking, these errors are catastrophic.
LEVIATHAN leverages:
- COMP-3 (Packed-Decimal) Arithmetic: COBOL’s native binary-coded decimal format that ensures 100% mathematical integrity for every cent processed.
- Batch Processing Power: Engineered to ingest and reconcile 50,000+ daily ledger entries in seconds.
- Immutable Auditing: Generates fixed-format reconciliation reports that meet strict mainframe audit standards.
- High-Precision Reconciliation: Automatic balancing of Credits (Inflow) vs Debits (Outflow).
- AML Anomaly Detection: Integrated logic to flag high-value suspicious transfers (Anti-Money Laundering compliance).
- Native Mainframe Performance: Compiled to native binary via GnuCOBOL for maximum execution speed.
- Streamlit Integration: A sophisticated web-based Command Center to monitor batch execution and audit logs.
To run this mainframe-grade engine on the cloud:
- Ensure
gnucobolis defined in yourpackages.txt. - Ensure
streamlitis defined in yourrequirements.txt. - The gateway will compile the
.cblsource into a native binary and execute the EOD (End-of-Day) routine.
Architected by UBAYDA ALİ to demonstrate the resilience of legacy financial cores within a modern cloud ecosystem.
- LinkedIn: Connect with me
- Portfolio: onws.net
- Live Engine: LEVIATHAN on Streamlit
- GitHub: Source Code
Precision is not an option; it is a fundamental requirement of the global financial system.