The Broken Reality of Lending Today

Over the past decade, banks and NBFCs have spent billions digitizing every step of the lending process: onboarding, KYC, documentation, disbursement, and collections.

But the one function that determines who gets credit and why — underwriting — has barely evolved.

Today's underwriters operate inside workflow tools that automate paperwork but not judgment. They face endless screens, data PDFs, bureau reports, and internal policies, yet decisions still rely on instinct, institutional habit, or “what we did last time.”

It's not a data problem anymore. It's a judgment problem, and it's costing institutions efficiency, consistency, and trust.

The institutional memory problem

Every credit institution, from banks to NBFCs, suffers from the same invisible disease: institutional amnesia. Every credit decision, every deviation note, and every risk memo contains invaluable intelligence: why a loan was approved, what exceptions were made, and what lessons were learned.

But these insights are trapped in silos: PDFs, Excel sheets, scattered MIS systems, and the minds of senior underwriters. When an experienced underwriter retires or switches jobs, that judgment doesn't transfer. It disappears. Each new loan becomes a reinvention of the wheel. Each branch applies policies differently. Each decision loses the context of the thousands before it.

This is why even the most advanced banks still struggle with inconsistent underwriting and credit drift across teams.

Judgment drift

Every loan decision is a representation of institutional philosophy. It reflects banks’ risk appetite amid real-world uncertainty. But when underwriting decisions vary widely between teams or geographies, it leads to judgment drift. Small inconsistencies compound over time into major portfolio distortions.

Two underwriters, given the same SME borrower, may make opposite calls, both backed by data. One focuses on collateral comfort, another on cashflow stability. Both are “right,” yet both expose the institution to unquantified bias and inconsistency. This drift doesn't show up in dashboards until it's too late: when delinquency trends rise, or regulators start asking questions about model governance.

Data isn’t the problem — reasoning is

In the last five years, India's lending ecosystem has seen an explosion of data access: Bureau APIs, GST feeds, account aggregators, alternate data sources, psychometric scoring, and more. But despite this abundance, decision-making hasn't become faster or more accurate.

Why? Because data volume ≠ underwriting intelligence.

Underwriters don't struggle with lack of data. They struggle with synthesizing it: making sense of multiple signals, remembering institutional context, and aligning decisions with credit policy frameworks. The gap isn't in information availability. It's in institutional reasoning — the process of turning data into decisions.

The lending industry doesn't need more dashboards or scoring APIs. It needs a new kind of intelligence — one that remembers, reasons, and aligns with institutional judgment.