Artificial intelligence (AI) is increasingly transforming the financial sector, particularly in improving the accuracy and efficiency of credit risk assessment. A community service project conducted by Rizqi Adhyka Kusumawati and Arief Darmawan from the Faculty of Business and Economics, Universitas Islam Indonesia (UII) developed an AI-based Credit Risk Rating (CRR) application to enhance financing risk management at BPRS UII, an Islamic rural bank in Yogyakarta. Published in June 2026, the project demonstrates how digital technology can support more objective, transparent, and consistent financing decisions while adhering to Islamic banking principles.
BPRS UII faces financing risks because creditworthiness assessments have traditionally depended heavily on analysts' subjective judgment. Manual evaluation processes may lead to inconsistent decisions and increase the possibility of misclassifying financing risks. Recognizing these challenges, the researchers designed an AI-assisted decision support system accompanied by a standardized Credit Risk Scoring training module to strengthen employees' understanding of financing risk management.
The community service program was implemented over six months, from October 2024 to March 2025, using a participatory and collaborative approach. The activities included field observations, needs assessment, application design, employee training, mentoring, and evaluation. Rather than replacing human decision-makers, the application was developed to assist financing analysts by providing structured and standardized risk assessments.
The AI-based Credit Risk Rating application was built using the Flutter framework and the Dart programming language. It incorporates two different assessment models based on customer segments:
- Retail customers are evaluated using a logic-gate model based on Debt Service Ratio (DSR), SLIK credit history, and employment stability.
- Corporate customers are assessed using an expert system based on the 5C principles—Character, Capacity, Capital, Collateral, and Condition—with weighted scoring to produce a comprehensive credit rating.
According to the project results, several important outcomes were achieved:
- A functional prototype of the Credit Risk Rating application was successfully developed.
- A standardized Credit Risk Scoring training module was prepared for BPRS UII employees.
- Staff members improved their understanding of financing risk indicators and rating interpretation.
- The institution established a foundation for digital transformation in financing risk management while maintaining prudential and transparency principles required in Islamic financial institutions.
The application classifies financing eligibility using transparent scoring rules. For corporate financing, ratings range from AAA (highly eligible with very low default risk) to BBB (eligible with additional monitoring) and D (not eligible due to inadequate financial strength or collateral). This explainable approach allows analysts to understand the reasoning behind every recommendation instead of relying on opaque "black-box" AI models.
During the mentoring sessions, employees were trained to use the application responsibly. The researchers emphasized that AI should serve as a decision support system, not the sole basis for financing approval. Final decisions should continue to consider internal policies, Sharia compliance, field verification, supporting documentation, and management approval. This approach minimizes the risk of algorithmic bias while ensuring accountability in financing decisions.
Evaluation of the program demonstrated improvements in several key areas, including standardized risk analysis, employee capacity building, technology adoption, and data awareness. Although the prototype performed successfully, the researchers recommended future development through historical data validation, integration with standard operating procedures (SOPs), implementation of data governance, periodic staff training, and the creation of an AI-powered early warning system for financing risk.
The authors conclude that successful digital transformation in Islamic banking requires a balance between technology, organizational processes, and human expertise. By combining explainable AI with sound risk management practices, the Credit Risk Rating application provides BPRS UII with a practical framework for improving financing quality while supporting transparency, fairness, and sustainable digital innovation in Islamic financial institutions.
Author Profiles
- Rizqi Adhyka Kusumawati - Universitas Islam Indonesia
- Arief Darmawan - Universitas Islam Indonesia
Research Source
Kusumawati, R. A., & Darmawan, A. (2026). Designing an AI-Based Credit Risk Rating and Data Analysis Application to Improve the Effectiveness of Financing Risk Management at BPRS UII. Jurnal Pengabdian Masyarakat Bestari (JPMB), 5(6), 483–496.
0 Komentar