Internal Credit Rating (Lending Hub): Replacing Hardcoded Logic with Zero-Code Agility

In partnership with Vietnam’s largest commercial bank by total assets on their comprehensive lending ecosystem reconstruction (Lending Hub), the Evotek team was entrusted with a pivotal mission: rearchitecting the Internal Credit Rating engine—the core component dictating credit approval quality and risk-adjusted pricing across both corporate and retail portfolios.

Operational Bottlenecks in Legacy Infrastructure

Over a decade in operation, the legacy credit evaluation system (CROMS) had accumulated severe structural friction well-known to both business analysts and engineering teams:

• Formula updates required source code alterations: Financial markets evolve continuously. However, whenever risk modelers sought to recalibrate a financial ratio weight, incorporate an industry-specific metric, or adjust a rating scale, they remained entirely reliant on IT developers to modify source code. Lengthy software development, testing, and deployment cycles created painful lags between strategic credit decisions and operational execution.
• Manual ingestion and cross-verification of financial statements: Customer information and financial statements were scattered across siloed departmental applications, forcing branch officers to spend excessive hours on manual data entry and multi-system verification, elevating the risk of input errors.
• Diverse customer segments with distinct risk profiles: Manufacturing enterprises, commercial trading firms, FDI companies, and individual micro-businesses possess fundamentally different financial profiles. The legacy system lacked dynamic mechanisms to automatically recognize and apply segment-tailored evaluation models.

Architectural Blueprint: Configuration-Driven Architecture & Zero-Code Rules Engine

Rather than continuing to hardcode business logic into backend repositories, Evotek’s software architects addressed the challenge with a Configuration-Driven engineering paradigm:

1. Empowering risk analysts through a Zero-Code Rules Engine: We engineered an isolated, zero-code business rules engine. Calculation formulas, evaluation thresholds, weight scorecards, and rating scales (AAA through D) are abstracted into configuration data. Risk modelers can configure, simulate, and deploy rating models (both Logistic Regression and Basel-compliant scorecards) directly through an administrative portal without backend source code intervention.
2. Schema-driven Dynamic UI: Built upon an Angular Monorepo architecture, appraisal forms and financial input grids are dynamically rendered from schema metadata. When a rating model introduces a new metric, the frontend forms automatically adapt in real time.
3. Intelligent enterprise data federation: The engine integrates directly with the Customer Information File (CIF) database and central financial repositories to auto-populate historical balance sheets, automatically classifying business sectors to recommend the most suitable scoring model.
4. Auditable and governed rating overrides: When senior credit committees require qualitative rating overrides, the platform enforces strict notch-migration boundaries while recording immutable audit trails to comply with banking regulatory inspections.

Measurable Business & Operational Impact

The combination of modern software architecture and deep banking domain expertise delivered concrete, quantifiable operational outcomes:

• Instant model adaptation and rapid deployment: Risk management departments now exercise complete autonomy over modeling experiments, calibration, and rollout without waiting for IT release cycles.
• Drastically reduced appraisal turnaround times: Front-office branches saw hours of repetitive paperwork eliminated through automated CIF extraction, flexible Excel bi-directional integration, and automated generation of standardized PDF credit rating memos.
• Granular Risk-Based Pricing: Real-time, dependable rating results feed directly into loan pricing engines, empowering the bank to institute competitive risk-based interest rates while safeguarding portfolio margins.

The initiative stands as a testament to Evotek’s engineering commitment: not merely modernizing software, but liberating front-line bank officers from manual workflows and granting true operational autonomy to enterprise risk management teams.