AI-Powered Alternative Credit Scoring System for MSMEs and Individual Borrowers in India

Ghanshyam Chand Yadav¹, Vatsalya Pandey², Kunal Mishra³

¹ Assistant Professor, Department of Commerce, Delhi College of Arts and Commerce, University of Delhi, New Delhi, India.

^2,3 Student, Department of Commerce, Delhi College of Arts and Commerce, University of Delhi, New Delhi, India. 

Abstract

Access to credit continues to be an endemic problem in India, especially concerning micro, small and medium enterprises (“MSMEs”) and the expanding segment of individual borrowers who function outside the traditional financial system. While gig workers, first-time borrowers and entrepreneurs operating in the informal sector are all actually creditworthy, they cannot be identified by lenders because their financial behaviour is not captured in the traditional credit bureau systems. To address this significant problem, this paper proposes an AI-based alternative credit scoring prototype for the Indian market, using digital financial indicators from Unified Payments Interface (“UPI”), Goods and Services Tax Network (“GSTN”) compliance data, and operational business metrics. The AI system has been independently validated using two separate Gradient Boosting classifiers with Isotonic Calibration, trained on 5,000 synthetic records of individual borrowers and MSMEs in India. The individual models yielded Area Under Curve (“AUC”) and Brier scores of 0.7650 and 0.1950, respectively, while MSME models yielded AUC and Brier scores of 0.7840 and 0.1840.

In addition to predictive accuracy, the system has integrated SHAP-proxy explainability for transparency at the borrower level, a three-tier probability of default decision-making framework, portfolio-level Expected Loss simulations using EL = PD x LGD x EAD, and a Disparate Impact Ratio governance dashboard. The portfolio level results indicated that the Expected Credit Loss was reduced by 51.7% over randomly selected accounts. All four design dimensions are aligned with the Reserve Bank of India FREE-AI framework, the Digital Lending Directions (2022) and the Digital Personal Data Protection Act (2023). The study addresses five documented research gaps in the Indian AI credit scoring literature and offers a replicable, regulatorily compliant blueprint for responsible AI deployment in India’s lending ecosystem.     

Keywords: Artificial Intelligence (AI), Credit Scoring, MSMEs  

References

1. Bazarbash, M. (2019) FinTech in Financial Inclusion: Machine Learning Applications in Assessing Credit Risk. IMF Working Paper No. 19/109, International Monetary Fund. https://www.imf.org/en/Publications/WP/Issues/2019/05/17/FinTech-in-Financial-Inclusion-Machine-Learning-Applications-in-Assessing-Credit-Risk-46883 
2. Berg, T., Burg, V., Gombović, A., et al. (2020) ‘On the rise of FinTechs: Credit scoring using digital footprints’, Review of Financial Studies, 33(7), pp. 2845-2897.
3. Brown, I. and Mues, C. (2012) ‘An Experimental comparison of classification algorithms for imbalanced credit scoring data sets’, Expert Systems with Applications, 39(3), pp. 3446-3453.
4. Deswal, S. and Pal, M. (2025) ‘Uncertainty estimation in predicting oxygenation by plunging jet aerators using probabilistic machine learning and conformal prediction’, International Journal of Technology, Health and Sustainability, 1(2), pp. 83-93. https://ijths.com/wp-content/uploads/2025/12/IJTHS-010230.pdf
5. Frost, J., Gambacorta, L., Huang, Y., et al. (2019) BigTech and the Changing Structure of Financial Intermediation. BIS Working Papers No. 779, Bank for International Settlements. https://www.bis.org/publ/work779.htm 
6. Gambacorta, L., Huang, Y., Qiu, J., et al. (2019) How do machine learning and non-traditional data affect credit scoring? BIS Working Paper No. 834, Bank for International Settlements. https://www.bis.org/publ/work834.htm 
7. GOI (2023) Digital Personal Data Protection Act, 2023. Ministry of Electronics and Information Technology, Government of India. https://www.meity.gov.in/content/digital-personal-data-protection-act-2023 
8. Hand, D.J. and Henley, W.E. (1997) ‘Statistical classification methods in consumer credit scoring’, Journal of the Royal Statistical Society: Series A, 160(3), pp. 523-541.
9. IFC (2017) MSME finance gap: Assessment of the shortfalls and opportunities in financing micro, small and medium enterprises in emerging markets. International Finance Corporation, World Bank Group, Washington, DC.
10. Jagtiani, J. and Lemieux, C. (2019) The Roles of alternative data and machine learning in FinTech lending, Federal Reserve Bank of Philadelphia Working Paper. https://doi.org/10.21799/frbp.wp.2018.15 
11. Khandani, A.E., Kim, A.J. and Lo, A.W. (2010) ‘Consumer credit risk models via machine learning algorithms’, Journal of Financial and Quantitative Analysis, 45(6), pp. 1639-1669.
12. Lessmann, S., Baesens, B., Seow, H., et al. (2015) ‘Benchmarking state-of-the-art classification algorithms for credit scoring’, European Journal of Operational Research, 247(1), pp. 124-136.
13. Lundberg, S.M. and Lee, S.I. (2017) ‘A unified approach to interpreting model predictions’, Advances in Neural Information Processing Systems (NeurIPS), 30, pp. 4765-4774.
14. Mehrabi, N., Morstatter, F., Saxena, N., et al. (2021) ‘A survey on bias and fairness in machine learning’, ACM Computing Surveys, 54(6), pp. 1-35.
15. MoMSME (2022) Annual Report 2021-22. Ministry of Micro, Small and Medium Enterprises, Government of India. https://msme.gov.in/sites/default/files/Annualrprt2021-22.pdf 
16. NABARD (2023) MSME Credit Demand and Supply Gap Assessment. National Bank for Agriculture and Rural Development. https://www.nabard.org/auth/writereaddata/tender/1003240118MSME%20Credit%20Demand%20and%20Supply%20Gap%20Assessment%202023.pdf 
17. RBI (2021) Master Direction — Account Aggregator Framework. Reserve Bank of India. https://rbi.org.in/Scripts/BS_ViewMasDirections.aspx?id=11284 
18. RBI (2022) Digital Lending Guidelines. Reserve Bank of India. https://rbi.org.in/Scripts/NotificationUser.aspx?Id=12345&Mode=0 
19. RBI (2023) Financial Inclusion Statistics. Reserve Bank of India. https://rbi.org.in/Scripts/AnnualPublications.aspx?head=Financial%20Inclusion%20Statistics 
20. RBI (2025) Framework for Responsible and Ethical Artificial Intelligence (FREE-AI). Reserve Bank of India. https://rbi.org.in/Scripts/PublicationReportDetails.aspx?UrlPage=&ID=1278 
21. SIDBI and TransUnion CIBIL (2022) MSME Pulse Report. Small Industries Development Bank of India. https://www.cibil.com/resources/msme-pulse-report 
22. Thomas, L., Crook, J. and Edelman, D. (2002) Credit Scoring and Its Applications. Philadelphia: SIAM.
23. WB (2017) Alternative Data Transforming SME Finance. World Bank.  https://documents.worldbank.org/en/publication/documents-reports/documentdetail/159151505845147395/alternative-data-transforming-sme-finance