India has built the most sophisticated digital financial infrastructure on the planet, featuring the Unified Payments Interface (UPI), account aggregator, goods and services tax (GST), Open Network for Digital Commerce (ONDC), Unified Lending Interface, among others; a real-time data spine connects over a billion people to the formal economy. And then a small business owner walks into a bank and gets asked for an audited balance sheet.

The balance sheet was designed for a borrower that most Indian MSMEs (micro, small, and medium enterprises) are not and never were. Built for large manufacturers with CFOs (chief financial officers) and structured annual disclosures, it was never meant to evaluate a family-run distributor in Surat or a logistics operator in Jaipur whose business is real, but whose paperwork is thin. India’s 63 million MSMEs, employing over 110 million people and generating nearly 30% of the gross domestic product (GDP), were never that borrowers.

When a lender looks at a small business through a balance sheet lens and sees nothing, that is not an accurate risk assessment. It is a good instrument applied to the wrong person. And it is costing India somewhere between Rs 20 and Rs 25 lakh crore in unmet MSME credit demand every year, an estimate the Reserve Bank of India (RBI)’s own expert committee put on record in 2019—and it has barely shifted since.

Here is what is maddening: India already has better instruments. It just has not made them the primary standard.

A GST filing tells you what a business sold, to whom, and when, across every quarter for years. The account aggregator pulls three years of live bank transaction data in under a minute. UPI histories, now running at over 16 billion transactions a month, reveal the daily economic rhythm of millions of small merchants who may have never been inside a formal credit system. The Open Credit Enablement Network (OCEN) is building the plumbing for lenders to plug into this data at scale. ONDC is creating a digital commerce layer where transaction histories become credit signals. ULI, piloted by RBI in 2023, is designed to let lenders access this entire stack frictionlessly.

This data does not approximate business health. It is business health, recorded in real time, far more honestly than any document assembled once a year for the purpose of securing a loan.

No other country has this combination at this scale. India accounts for 49% of all global real-time payment transactions. The US does not have this stack. China does not have it the way India does. This infrastructure was built here, at enormous public cost and political will, and got a billion people using it. Yet only 14% of Indian MSMEs have access to formal credit today. The data is already flowing. What lags is the willingness to treat it as first-class evidence of creditworthiness rather than a supplementary comfort check.

Consider a business owner who has been filing GST for six consecutive years, has never missed a vendor payment, and has a transaction history that would put most salaried borrowers to shame. She gets declined for a term loan because her fixed assets fall short of collateral thresholds. The bank is not being unreasonable within its own framework. The framework is simply built for someone else.

Rewrite the framework and you rewrite the outcome.

RBI’s push toward cash flow-based lending is directionally right. The rails are live. AI-driven underwriting models that read these signals already exist and are already working inside Indian lenders, delivering higher approvals with lower defaults, because they are reading real signals instead of lagging documents. The credit officer is not the obstacle. Given richer inputs, good credit officers make better decisions. What limits their judgement today is the requirement to work backwards from inadequate information.

India has a once-in-a-generation opportunity to become the first major economy to build a credit system genuinely designed around how small businesses actually operate, not around how traditional banking systems assume they should. The infrastructure investment has been made. The rails are live. The models are ready.

What still needs to change is what we accept as proof that a business deserves credit and an ability to generate the right intelligence from this data using new-age underwriting methodologies, often using AI.

Until then, 63 million businesses will keep queuing at a counter that was built for someone who looks nothing like them.

Joydip Gupta is APAC Head at Scienaptic AI, an IIT Delhi alumnus and ex-McKinsey consultant with over two decades in credit technology and financial inclusion across India and Asia.

Chandan Pal is Chief Marketing Officer at Scienaptic AI and an IIM Indore alumnus. Views are personal.