Transforming Small Businesses 2025

USE CASE ILLUSTRATION 2 AI agent-enabled credit access for MSMEs An Indian financial technology (fintech) company is trying to bridge the lending gap in the MSME ecosystem by providing loans to micro-merchants and small businesses. The current lending process consists of three stages. The first is the application stage, in which MSMEs submit the required documentation for the loan application and credit appraisal. This is the most challenging stage for MSMEs as they lack structured financial and transactions data. The second is the credit appraisal stage, which involves quantitative and qualitative analysis to assess the MSME’s ability and intent to pay. The third is post-disbursement risk management, where collections data is analysed to assess ongoing changes in the risk profile of the MSME. The current lending process is full of challenges, for both the lender and the MSME. MSMEs are often unaware of government schemes, loan eligibility and institutions that can offer them loans. These organizations rely on direct selling agents (DSAs), which connect them with financial institutions and help them in the loan process. The DSAs charge a commission for this service, which increases the cost of credit for MSMEs. On the other hand, lenders still rely on traditional methods such as physical visits to analyse creditworthiness. The use of AI agents can simplify the process for both parties. An illustrative example of AI an agent-enabled lending process for MSMEs is shown below. Note: AOP = annual operating plan; GSTIN = Goods and Service Tax Network number; KYC = know your customer; GST = goods-and-services tax. Source: World Economic ForumUsing AI agents to simplify credit for MSMEs Registries available at government ministriesDynamic or real-time MSME data Supporting registries MSME schemes registry Financial institutions registryCore registries Udyam registration number GSTIN Georefer- enced location Product registryDynamic MSME data Financial statements Bank statements Production data AOP dataAgent A: Loan advisor Analyses MSME data, government schemes and financial institutions data (interest rates, eligibility, etc.) to recommend loan type and lenderAgent B: Documentation Collates KYC, financial and bank data and prepares loan application Agent 1: KYC Uses Udyam ID and GST for e-KYC of applicant Alternative data sources Order book Supply chain Customer reviewsAgent 2: Risk analyser Identifies industry- specific growth drivers and risksAgent 3: Credit appraiser Combines financial and bank data along with alternate data sources for comprehensive credit appraisalAgent 4: Underwriter Finalizes loan terms, interest rates and tenure based on risk assessmentAgent 5: Monitoring Tracks MSME and industry performance on growth drivers and risk signalsAgent C: Data coordinator Provides alternate data requested by credit appraisal agentMSMEAgent D: Feedback Analyses financial health and suggests methods of improvementMSME meta agent Lender meta agentFinancial institutionsGovernment schemes Investment plansCredit informationProduction and sales targetsBank statementsFinancial statementsMSME data Loan disbursedMSME financial manager approves Transforming Small Businesses: An AI Playbook for India’s SMEs 36