Architecting India’s Credit DPI For High-Growth MSMEs

Architecting India’s Credit DPI For High-Growth MSMEs

By Pranay Bhargava and Vijay Mahajan Vijay Mahajan is Founder, BASIX Social Enterprise Group and Director, Rajiv Gandhi Institute for Contemporary Studies. Pranay Bhargava is Founder, Micro-Equity Fund, Impact Micro Ventures and Enable Livelihoods Foundation.

Abstract

India’s Digital Public Infrastructure (DPI) — including Unified Lending Interface (ULI), Open Credit Enablement Network (OCEN), Account Aggregator (AA), Unified Payments Interface (UPI) and Open Network for Digital Commerce (ONDC) — has successfully enabled real-time digital lending but primarily serves Own Account Enterprises (OAEs). The existing DPI architecture neglects the critical “missing middle”: Hired Worker Enterprises (HWEs) and Extraordinary Aspiration HWEs (αHWEs), which offer the highest potential for growth, scalability, and job creation in local economy.

To fully realize DPI’s transformative potential, India must shift from rigid, EMI-based lending toward flexible, performance-linked, and risk-sharing capital solutions. This paper identifies key strategic shifts to achieve this transition: explicitly targeting HWEs and αHWEs as primary beneficiaries; embedding flexible finance (e.g. revenue share & profit share) and micro-equity into ULI, OCEN, UPI, and ONDC; transforming AA from a data aggregator into an intelligent risk infrastructure; and leveraging RBIH-led sandboxes to strengthen and institutionalize flexible financing models.

With the foundational infrastructure already in place, India has a golden opportunity to unlock ₹15–25 lakh crore of flexible financing. Such an evolution can catalyze 10 crore new jobs, promote enterprise growth, and spark a revolution as powerful as microfinance movement which happened three decades ago.

Key Words: Digital Public Infrastructure (DPI) India, Flexible Finance, Micro-equity Funding, MSME Financing India, Cashflow-based financing, Unified Lending Interface (ULI), Open Credit Enablement Network (OCEN), Account Aggregator (AA), Unified Payments Interface (UPI), Open Network for Digital Commerce (ONDC), High-growth Aspirational Entrepreneurs, Reserve Bank Innovation Hub (RBIH), Agentic AI

Table of Contents

1 India Needs Jobs — and HWEs and αHWEs Are the Key

1.1 Own Account Enterprises (OAEs)

1.2 Hired Worker Enterprises (HWEs)

1.3 Extraordinary Aspiration HWEs (αHWEs)

1.4 Economic Impact: Multiplier and Network Effects

1.5 Policy Gap and Imperative to Shift Focus

2 India’s Current DPI for Credit

2.1 Current System Prioritizes OAEs and Offers Inflexible Loan Products

2.2 True Potential Lies in Serving Growth-Oriented HWEs and αHWEs

3 How Must Credit DPI Be Designed to Serve HWEs and αHWEs

3.1 ULI and OCEN Must Introduce Flexible Finance and Micro-Equity

3.2 Leverage GSTN and ONDC Data for Offering Flexible Finance and Micro-Equity

3.3 Transform the Account Aggregator (AA) into an Intelligent Risk Infrastructure

4 Additional Features Needed for a Robust Credit DPI

4.1 Capture a Holistic Enterprise Dataset

4.2 Use AI to Automate the Enterprise Credit Lifecycle

4.3 Transform UPI into an Automated Flexible Repayment Infrastructure

5 Call to Action

5.1 Launch RBIH-Led, OCEN-Enabled Sandbox Pilots

5.2 RBIH and OCEN to Co-Develop an Agentic AI Layer for Enterprise Financing

5.3 Policy Leadership by the Ministry of Finance and the Reserve Bank of India

1 India Needs Jobs — and HWEs and αHWEs Are the Key[1]

India’s unincorporated sector employs over 90% of the non-farm labour force, making it a critical engine of economic growth. Entrepreneurs in India’s unincorporated sector broadly fall into three categories, differentiated by their scale, ambition, and economic impact:

1.1 Own Account Enterprises (OAEs)

Share: Approximately 86% (6.34 crore out of total 7.34 crore) of enterprises.

Employing 66% of workers (7.65 crore out of 12.06 crore workers).

Scale & Scope:

Employ 1–2 individuals, generally the entrepreneur and family member.

Typical investment: below ₹2.5 lakh.

Annual revenue around ₹2.5 lakh per annum.

Activities & Market:

Focused on hyper-local markets at village or panchayat level.

Undertake basic value-added activities such as pre-processing, trading, or retail.

Finance Needs:

Adequately served by existing microfinance and SHG ecosystems.

1.2 Hired Worker Enterprises (HWEs)

Share: Approximately 13% (1.00 crore out of total 7.34 crore) of enterprises.

Employing 34% of workers (4.41 crore out of 12.06 crore workers).

Scale & Scope:

Employ around 4–5 workers.

Require investments of ₹5–25 lakh.

Generate annual revenues around ₹18 lakh per annum.

Activities & Market:

Operate at block-level markets and cater to external markets.

Engage in medium-level value-added activities like primary processing (sorting, grading, packaging), distribution, and localized aggregation.

Finance Needs:

Require flexible financing options for fixed capital.

Need traditional debt for working capital.

1.3 Extraordinary Aspiration HWEs (αHWEs)

Share: Approximately 1% (about 7.5 lakh out of 7.34 crore) of enterprises.

Employing about 12.4% (about 1.5 crore out of 12.06 crore workers).

Scale & Scope:

Employ over 20 individuals on an average (range 10 -100).

Require investment of ₹1 crore or more.

Achieve annual revenues ₹ 2–5 crore.

Activities & Market:

Serve district, state, or regional markets, including metropolitan areas.

Perform higher-order value-added activities such as secondary processing, branding, and large-scale aggregation or distribution.

Finance Needs:

Require micro-equity financing for fixed capital.

Require customized, cashflow-linked loans for working capital.

1.4 Economic Impact: Multiplier and Network Effects

αHWEs and HWEs function as critical drivers of local and regional economies. αHWEs, akin to “micro-unicorns,” act as regional anchors or super-aggregators, creating substantial market demand. This demand trickles down, providing markets to multiple HWEs, who in turn provide markets to even larger number of OAEs. Thus, a powerful multiplier effect emerges from αHWEs, setting off a dynamic circulation of goods, services, and capital within the community network.

Additionally, the agglomeration of αHWEs, HWEs, and OAEs generates significant network effects. Each enterprise segment benefits from mutual interactions: αHWEs offer foundational incubation support, finance, and ready markets to HWEs, while HWEs assure steady supply chains for αHWEs. Similarly, HWEs and OAEs share a comparable symbiotic relationship. This creates an ecosystem where the strength of the network is greater than the sum of its individual enterprises, thereby incentivizing everyone to join and stay within the network.

1.5 Policy Gap and Imperative to Shift Focus

Despite their high potential for driving regional growth, αHWEs — and to some extent HWEs — have remained largely overlooked by policymakers, mainly due to their smaller numbers and systemic incubation and financing bottlenecks that prevent them from thriving. Consequently, policy efforts stay heavily focused on OAEs, leaving αHWEs and HWEs to navigate challenges on their own.

The need of the hour is to shift policy, efforts, and resources earnestly and dedicatedly toward αHWEs and HWEs if India is to achieve an extraordinary increase in employment. The key to enabling this crucial shift lies in redirecting the country’s credit DPI toward αHWEs and HWEs — specifically toward what they need to thrive: flexible financing and micro-equity financing.

2 India’s Current DPI for Credit

India’s Digital Public Infrastructure (DPI) began with the UIDAI and Aadhaar cards, followed by the Unified Payments Interface (UPI). The DPI for Credit — comprising the OCEN and the ULI — is a more recent development. It holds the promise of revolutionizing access to finance, particularly for India’s large base of micro-enterprises, estimated at 7.34 crore as per ASUSE 2023–24.

So far, however, the focus has been primarily on enabling instant, small-ticket, short-duration loans tailored for nano enterprises — the OAEs[2]. Within this framework, OCEN enables interoperable loan offers, while ULI facilitates rapid documentation and onboarding[3].

2.1 Current System Prioritizes OAEs and Offers Inflexible Loan Products

Both ULI and OCEN remain optimized for rigid, EMI-based products such as invoice financing[4] and Pradhan Mantri Mudra Yojana (PMMY) loans, which address only a part of the financial needs of nano enterprises (OAEs).

As evidence, the average loan size under PMMY between 2016 and 2025 has been ₹55,623. Even more telling is that 83.4% of all PMMY loans have gone to the below-₹50,000 Shishu category, where the average loan size was ₹27,057. Only 2.0% of the loans have been extended to the Kishore category (typically representing hired worker enterprises), with an average disbursed loan size of ₹654,509. It is important to note that this figure reflects the disbursement value; during the lifetime of the loan, the average outstanding amount tends to be about half of the disbursed amount.

While focusing on OAEs has helped advance financial inclusion for enterprises at the base of the pyramid, it has also overlooked the much bigger credit challenge faced by India’s most transformative businesses — HWEs and αHWEs. These growth-oriented businesses require need-based, larger-ticket financing (₹5–25 lakh) for longer durations (3–5 years), structured around their cashflow variability and performance-linked repayment capacity. What they need is not instant microcredit, but larger, flexible finance and micro-equity solutions — needs that current DPI flows do not yet adequately support.

2.2 True Potential Lies in Serving Growth-Oriented HWEs and αHWEs

Ironically, while current DPI use cases cater largely to OAEs — who often lack digital records, operate informally, and show limited repayment capacity — DPI is structurally better suited for HWEs and αHWEs. These enterprises, although underserved by traditional finance due to perceived risks like revenue seasonality and informal nature of business, are relatively more digitally active[5] and can generate the kinds of transaction data (via GST, UPI, ONDC, AA) that DPI is designed to leverage.

3 How Must Credit DPI Be Designed to Serve HWEs and αHWEs

The Credit DPI is still relatively new and can be designed to serve a more promising segment — both in terms of growth and employment potential — of the Indian enterprise sector: the HWEs and αHWEs. The upcoming credit DPI frameworks, such as ULI and OCEN, must treat HWEs and αHWEs as priority, given their economic multiplier effect and digital readiness.

3.1 ULI and OCEN Must Introduce Flexible Finance and Micro-Equity

The upcoming credit DPI platforms must evolve rapidly to support new loan product templates that reflect the diverse capital needs and repayment realities of these enterprises. Key features should include:

· Demand-based financing: Larger loans with longer tenures (e.g., ₹5–25 lakh for 3–5 years).

· Flexible Financing with Repayments Linked to Cashflows (FFRC):

o Revenue-based financing (e.g., 5% of monthly gross sales),

o Profit-sharing structures (e.g., 20% of monthly profit),

o Dynamic tenure loans (e.g., repayment until a target multiple is reached),

o Deferred-start loans with ballooning repayments (e.g., “repayment begins after monthly sales cross ₹X” — ideal for seasonally ramping businesses).

· Micro-equity or equity-like participatory capital:

o Equity-style investments such as ₹15–25 lakh for a five-year term,

o Innovative instruments like micro-convertibles or SAFE[6] notes that align financier returns with enterprise performance without diluting ownership prematurely.

3.2 Leverage GSTN and ONDC Data for Offering Flexible Finance and Micro-Equity

The transactional data from the GST Network (like e-way bills and GST invoices) and ONDC shouldn’t just be used to support commerce — it should also power a data-driven embedded finance layer. This layer can offer flexible finance and micro-equity products, especially for HWEs and αHWEs who trade actively on these platforms. Loan or equity offers can be triggered automatically at the time of each transaction, based on real-time business activity. Key enablers can include:

· Access to flexible finance through the seller dashboard, based on real-time performance.

· Use of live transaction data for cashflow-based underwriting[7]. Pre-approved, flexible financial products offered through seamless integration with OCEN and AA.

For example, if a seller receives a big order, the platform can immediately offer them a loan to help fulfill it. Once the seller delivers the order and earns money, a small part of that payment can be automatically deducted to repay the loan. This way, sellers get the support they need exactly when they need it, without any paperwork or stress about separate repayments.

3.3 Transform the Account Aggregator (AA) into an Intelligent Risk Infrastructure

The AA framework holds immense potential to address a persistent barrier in MSME financing: the inability to assess creditworthiness beyond traditional financial records. While AA currently facilitates access to structured data — such as bank transactions, GST returns, and UPI trails[8] — its value remains underutilized for high-growth enterprises like HWEs and αHWEs.

These enterprises often remain excluded from formal credit due to the non-linear nature of their financial behaviour, including:

· Seasonal and volatile revenues, especially during startup and scale-up phases.

· Irregular but responsible repayment patterns that do not align with rigid credit bureau norms such as “Days Past Due (DPD).”

To address this, AA must evolve from a passive data pipe to an intelligent, AI-powered risk infrastructure — capable of building a real-time, multidimensional profile of enterprise viability and enabling performance-linked finance.

4 Additional Features Needed for a Robust Credit DPI

Beyond enabling ULI and OCEN to offer Flexible Finance and Micro-Equity products, and using GSTN and ONDC data to support these offerings, the Account Aggregator (AA) system must also be upgraded into an Intelligent Risk Infrastructure[9]. In addition, several completely new features will need to be added to strengthen the Credit DPI.

4.1 Capture a Holistic Enterprise Dataset

To reflect the true viability of HWEs and αHWEs, AA integrations must go beyond current data sources to include:

· Comprehensive Business Cashflows: Including geolocation-based market demand, asset base (machinery, inventory, real estate), recurring operational costs (utilities, rent, marketing), and labour payments — critical for assessing unit economics and working capital dynamics.

· Expanded Repayment Histories: Captured not just via traditional credit bureaus but from digital sources such as accounting platforms, e-commerce marketplaces, NBFCs, and micro-lenders.

· Psychometric Indicators: Digitally assessed traits such as aspiration, self-efficacy, risk appetite, and perseverance — highly predictive of entrepreneurial success in informal markets.

· Alternative Data Streams: POS usage, RFID scans, wallet app activity, and even CCTV-based footfall analytics — already used in consumer fintech — can offer real-time insight into business performance.

4.2 Use AI to Automate the Enterprise Credit Lifecycle

To make FFRC and Micro-Equity viable at scale, India’s DPI stack must support dynamic, AI-based credit intelligence throughout the lending lifecycle:

Lifecycle Stage: Sourcing At the sourcing stage, AI classifies borrowers into three risk categories using psychometric profiles, behavioral insights, and cashflow data. Borrowers identified as high-risk (“Red”) are directed to human underwriters, medium-risk (“Orange”) cases are routed through AI-assisted workflows, and low-risk (“Green”) cases are automatically sanctioned. This triaging system helps reduce operational costs and accelerates loan disbursals.

Lifecycle Stage: Monitoring During the monitoring phase, real-time enterprise health is tracked through dynamic Management Information Systems (MIS) and behavioral signals. This enables the early detection of financial stress or rapid growth, allowing interventions like pausing repayments or offering scale-linked support.

Lifecycle Stage: Collections Collections are managed through GenAI-powered nudges delivered in vernacular languages. These nudges are timed with the borrower’s business cash cycles, making them more empathetic and effective. This approach preserves the entrepreneur’s goodwill while reducing non-performing assets (NPAs) and collection costs — especially important for flexible, non-EMI-based financial products.

Strategic Value Propositions:

· Manual management of FFRC and micro-equity is impractical at scale. Embedded AI agents can make flexible models operationally viable for mass-market lenders.

· Automated sourcing, monitoring, and collections reduce the cost-to-serve — crucial for offering flexible financing products for sub-₹10 lakh loan sizes.

· Predictive scoring based on alternative and behavioural data unlocks credit for entrepreneurs who were previously considered ‘unlendable’ or had ‘impaired credit bureau’ records but are now ready for growth.

4.3 Transform UPI into an Automated Flexible Repayment Infrastructure

A defining feature of flexible finance isn’t just how credit is assessed — it’s how it’s recovered. To operationalize revenue-share, profit-linked, and deferred repayment models at scale, India must establish a standardized, API-based repayment infrastructure within its DPI.

Currently, all repayment rails — eNACH, UPI Autopay, standing instructions — are optimized for fixed EMIs[10]. They are ill-suited to handle dynamic or performance-linked obligations. This bottleneck threatens the viability of non-EMI credit models such as FFRC and Micro-Equity.

Therefore, DPI must support a programmable repayment architecture that is plug-and-play for lenders and seamless for borrowers.

Core Components of the Repayment Architecture

The Repayment API Stack, integrated with UPI, is designed to automate flexible repayments by computing, deducting, and distributing payments as a fixed percentage of a business’s revenue or profit. These APIs pull real-time data from Account Aggregator, GST, ONDC, or UPI merchant transactions. For instance, a repayment instruction could specify, “Deduct 5% of all UPI merchant credits until ₹X is recovered.” This automation enables true cash-flow-linked lending, allowing financial products to dynamically adjust based on actual business performance.

The Trigger Protocols layer introduces event-based and time-based repayment triggers. Lenders or platforms can define conditions such as “start deduction only when monthly revenue exceeds ₹1 lakh,” “EBITDA is above ₹5,000,” or “cash in bank surpasses ₹50,000.” Seasonal triggers, like initiating repayment only during peak business months, are also possible. These intelligent triggers ensure that repayments begin only when the business is financially viable, significantly reducing the risk of premature defaults.

Finally, the Lender Interface Controls offer configurable dashboards for lenders to adjust repayment terms without human intervention. These controls enable dynamic changes in revenue-share percentages, allow milestone-based top-up disbursals, and support exit options like share buybacks. This programmable flexibility builds trust with borrowers and enforces repayment discipline, all while reducing administrative overhead for lenders.

**#* A promising implementation pathway for flexible finance can be automated revenue-share repayment, directly linked to the merchant’s sales. For instance, repayments can be auto-deducted via a UPI merchant QR code tied to a special account[[11]](#_ftn11)*, with a fixed percentage of each sale — whether online or offline — channelled to the lender. This ensures that repayments are fully aligned with business cash flows: when sales dip, repayments decline; when sales surge, the loan is paid off faster.

While UPI 2.0 and e-mandates already make this technically possible, what’s still missing is a common, open system built into India’s digital public infrastructure. ULI and OCEN can make this happen by creating a Revenue-Share Loan API, where a simple field like ‘repayment rate’ sets what percentage of sales will be deducted, and connected systems automatically pull live sales data from platforms like Account Aggregator, UPI, GST, or ONDC.

This system can be made even smarter by using automated contracts. For example, if a business’s GST sales are ₹X in a month, the bank can automatically deduct a small percentage (X × repayment rate) from their account using e-NACH. This kind of fully automatic repayment setup can first be tested in an RBIH sandbox, where important results like default rates, recovery costs, and how much money is recovered for every ₹ of sales are carefully tracked. Once it’s standardized within India’s digital public infrastructure (DPI), revenue-based lending will become plug-and-play — making it much easier for high-growth MSMEs to access flexible capital at scale.

5 Call to Action

5.1 Launch RBIH-Led, OCEN-Enabled Sandbox Pilots

To make Flexible Finance and Micro-Equity a permanent part of the system for HWEs and αHWEs — the Reserve Bank Innovation Hub (RBIH) must lead a new wave of sandbox pilots focused specifically on these models. Earlier sandbox pilots mainly tested digital delivery of traditional loans. Now, it’s critical to test capital products built for non-linear growth, unpredictable cashflows, and performance-linked repayments — all key features of HWEs and αHWEs.

As the tech backbone of India’s public credit infrastructure (e.g., ULI), RBIH is uniquely positioned to design and oversee such innovation cohorts. These pilots should span key sectors like agri-processing (across primary to tertiary levels), non-farm manufacturing, and handcrafted industries (handlooms, powerlooms, artisan clusters). Importantly, pilots should target two enterprise life stages:

· Startup and high-growth phase (post product-market fit, pre-breakeven),

· Moderate-growth phase (EBITDA-positive or post-breakeven).

Key Design Features of the Proposed RBIH Sandbox

The Sandbox Layer is critical for transitioning flexible finance and micro-equity from isolated pilots to a mainstream component of India’s financial system. At its foundation, Regulatory Approvals must be secured to legitimize instruments like micro-equity and to revise existing regulatory frameworks — specifically, changes in Days Past Due (DPD) norms and the Income Recognition and Asset Classification (IRAC) rules — to accommodate cashflow-based repayments. These changes are essential to create the legal and supervisory space for non-EMI lending products to operate sustainably at scale.

Next, updated Credit Norms must allow for the experimentation with new financial structures, including revenue-share and profit-share contracts, deferred repayment models, and performance-linked pricing mechanisms. These innovations enable a more adaptive credit system that aligns repayment obligations with actual business performance, as opposed to rigid, time-bound EMIs.

The Full-Stack Testing component ensures that the entire digital public infrastructure (ULI, OCEN, Account Aggregator, ONDC, and UPI) can support these new models in practice. This holistic testing verifies that the different DPI layers can interoperate smoothly to facilitate flexible finance products end-to-end — from origination to repayment.

Furthermore, the sandbox must include Agentic Lifecycle Testing, where AI-powered workflows are piloted across key functions. This includes automated Red/Orange/Green risk classification, real-time MIS dashboards, and vernacular GenAI bots that deliver nudges and support to borrowers. These capabilities are crucial to demonstrate that flexible finance can be operationally viable and low-cost, particularly for loans under ₹10 lakh.

Finally, a robust Measurement Framework must be put in place to evaluate the sandbox’s effectiveness. Key indicators include time to sanction for low-risk (Green) borrowers, the cost of human underwriting for Red and Orange cases, NPA risk under modified DPD definitions, and the amount recovered per ₹100 lent. These metrics are vital for building an evidence-backed case for policy reform and mainstream adoption of flexible finance instruments.

From Proof-of-Concept to Systemic Scale

Once the model is validated, RBIH can recommend formal upgrades to DPI protocols like ULI and OCEN to natively support flexible repayment flows and equity-like instruments. This would pave the way for RBI to officially approve revenue-based financing pilots, with the right borrower protection safeguards in place.

5.2 RBIH and OCEN to Co-Develop an Agentic AI Layer for Enterprise Financing

To scale flexible finance for India’s most dynamic but underserved enterprises — HWEs and αHWEs — India needs more than robust data rails and upgraded credit protocols. It needs a national Agentic[12] AI layer: an always-on, multilingual, low-cost virtual advisor embedded across the credit DPI stack (ULI, OCEN, AA, ONDC).

This Agentic AI won’t just be an automation tool — it will serve as a full-lifecycle finance and incubation partner, acting like a virtual loan officer, business coach, personal growth guide, and BDS provider for millions of first-generation and growth-stage entrepreneurs.

Core Agentic AI Functions Across the Enterprise Lifecycle

At the Sourcing stage, Agentic AI pre-screens potential borrowers using data from Account Aggregators (AA), GST filings, ONDC transaction history, and psychometric assessments. It classifies borrowers into Red, Orange, or Green risk bands. This risk-based routing significantly reduces underwriting costs by directing human attention only to higher-risk cases, while efficiently identifying enterprises with strong growth potential.

During the Foundational Support (Pre-Onboarding) phase, the AI delivers customized Entrepreneurship Development Programs (EDPs) grounded in David McClelland’s Achievement Motivation Theory. It assesses the entrepreneur’s mindset, motivation, and readiness for growth. This step is crucial in bridging the foundational mindset gap and selecting only those entrepreneurs who are psychologically and behaviorally equipped to absorb finance and scale their ventures.

In the Onboarding & Education stage, AI-powered chatbots or voicebots explain the terms of revenue-share finance, profit-sharing logic, and micro-equity structures in vernacular languages. This ensures entrepreneurs understand the flexible finance models, enhancing trust, financial literacy, and informed consent — key enablers of responsible borrowing and long-term engagement.

As part of Business Development Services (BDS) Enablement, the Agentic AI provides personalized guidance in areas such as branding, digital onboarding, regulatory compliance, and customer management. It also connects entrepreneurs to shared BDS platforms. Since the success of flexible finance and micro-equity depends heavily on enterprise growth, enabling access to BDS is essential to maximize returns for both entrepreneur and financier.

In the Relationship Building & Monitoring phase, the AI monitors real-time business cashflows, behavioral patterns, and seasonality. It assesses business health, provides early warnings for financial stress, and even supports mental well-being. Additionally, it facilitates peer networking and local linkage creation, helping entrepreneurs embed themselves in supportive ecosystems that drive sustained growth.

For Repayment & Collections, the AI dynamically adjusts repayment schedules using UPI-based APIs, ensuring deductions match actual revenue flows. Rather than relying on rigid EMIs or legal pressure, the system uses personalized, vernacular nudges to prompt repayment. This humane approach maintains goodwill, improves repayment behavior, and reduces costs — especially vital for non-EMI products like revenue-sharing contracts #.

Finally, under Product Discovery & Upsell, the Agentic AI recommends tailored financial products — such as top-up flexible loans, insurance, micro-investments, or micro-equity options — based on enterprise milestones and behavior. This builds long-term financial partnerships that support enterprise expansion while deepening financial inclusion.

# For example, instead of sending a legal notice after a delay, AI could simply ask: ‘How is your business doing this month? Would you like a grace period?’ If the borrower agrees, the system could automatically adjust the repayment timeline — preserving trust and reducing the risk of default. Over time, this builds ‘relationship equity,’ which becomes as valuable as credit history for long-term financial inclusion.

Strategic Value for India’s DPI

· Massive Cost Efficiency: Enables the credit DPI to manage ₹10–25 lakh flexible finance at minimal per-user cost — impossible under manual models.

· Possibility of Re-Inclusion: Onboards MSMEs excluded due to damaged credit scores[13], using behavioural and alternative data for eligibility assessment.

· Adaptive Product Innovation: Real-time AI feedback sharpens credit product design, helping DPI refine FFRC and micro-equity offerings as the market evolves.

· Trust-Led Digital Interface: Builds a human-like relationship between entrepreneurs and the financial system — without dependence on physical bank branches.

To operationalize this system, the RBIH may take the lead to develop a national open-source Agentic AI toolkit hosted within the ULI/OCEN sandbox layer under the credit DPI governance. It may have LLM APIs, multilingual plug-and-play integrations for banks and NBFCs and embedded compliance checks to ensure transparent AI operations.

5.3 Policy Leadership by the Ministry of Finance and the Reserve Bank of India

India is doing well on economic growth, but employment remains a concern, along with the broader challenge of ensuring inclusive growth. The MSME sector represents the next great hope for employment generation — but within it, it is not the 6.34 crore OAEs that will drive job creation. It is the 1.0 crore HWEs that, given adequate access to finance for startup, growth, and navigating occasional shocks, can generate crores of new jobs. However, the present banking sector must significantly change its approach to enterprise lending.

As we have stated earlier, the level of transformation required — in policies, products, and processes — is similar to the shift that transformed old-generation IRDP poverty alleviation loans, which had repayment rates of only 18–20%, into the new-generation Self-Help Group (SHG) lending model, where repayment norms of 95–99% became the standard. Achieving this level of transformation requires strong policy leadership, a task worthy of the Ministry of Finance and the Reserve Bank of India.

The encouraging fact is that both the demand and the building blocks for breakthrough solutions already exist. New credit worth ₹15–20 lakh crore and potentially over 10 crore new jobs can be unlocked through decisive policy action.

References and Foot Notes —

[1] Quantitative data on enterprise distribution, employment, investment, and annual revenue is sourced from the Annual Survey of Unincorporated Sector Enterprises in India (ASUSE 2023–24), conducted by the NSSO, GoI. https://mospi.gov.in/sites/default/files/publication_reports/ASUSE_2023_24_Full_Report-L.pdf. Qualitative insights on enterprise activities and market behavior are based on the authors’ field-level experience across multiple states. Insights regarding the financing needs of different enterprise segments are derived from the policy paper: Mahajan, Vijay and Bhargava, Pranay, “SME Financing — How to Bridge the Persistent Demand Supply Gap?” (February 17, 2025), available at SSRN: https://ssrn.com/abstract=5141173 or http://dx.doi.org/10.2139/ssrn.5141173.

[2] Initiatives like PM SVANidhi, GST Sahay, and PSB Loans in 59 Minutes serve OAEs (e.g., street vendors, sole proprietors). UPI, GSTN, and AA integrations are being used primarily for small-ticket lending pilots through fintechs focused on nano and micro enterprises.

[3] ULI is still in prototype/pilot stage and is being tested for high-frequency, low-ticket lending scenarios (e.g., small traders, retailers, nano enterprises).

[4] OCEN 1.0 templates support standard EMI-linked products (e.g., short-tenure working capital loans, invoice discounting). While these small-ticket digital loans are supported in real-time through pilots like GeM SAHAY, flexible finance is not yet operational.

[5] As per ASUSE 2023–24, HWEs show higher formalization (47.8% licensed, 23.4% registered), better digital adoption (24.3% computer use, 58.5% internet access), and improved financial practices (5.7% maintain audited accounts).

[6] This may involve providing capital upfront in exchange for the right to receive equity at a future valuation event. This allows investor to participate in potential upside without requiring immediate valuation negotiations or complex structuring, making it particularly suitable for micro enterprises where conventional equity deals may be premature or burdensome.

[7] ONDC logs structured sales metadata — such as volume, frequency, price, and fulfillment — which can be leveraged for cashflow modeling. These transaction trails, when combined with data from the Account Aggregator and GSTN, can enable robust, DPI-aligned credit scoring.

[8] As of March 2025, AA networks allow consent-based sharing of data from banks, NBFCs, GSTN, TSPs, and some investment and insurance repositories. It’s live through the Sahamati ecosystem and being scaled by banks, fintechs, and NBFCs.

[9] AA remains a data pipe per RBI rules. Intelligence must be built on top by market participants, not by AA itself.

[10] UPI AutoPay supports recurring mandates, but flexible repayment like revenue-share needs a programmable repayment stack which does not yet exist natively in DPI.

[11] While UPI 2.0 supports e-mandates and recurring debit instructions — allowing lenders to collect fixed or pre-agreed payments from merchants, with mandates tied to specific accounts and set to recur (e.g., daily, weekly) — UPI APIs do not yet support native revenue-share logic, such as repayment_rate fields or dynamic deductions based on a percentage of sales.

[12] Agentic AI is a type of artificial intelligence that focuses on autonomous systems capable of making decisions and performing tasks independently, without human intervention. These systems can analyze data, set goals, and take actions, often adapting to changing environments and learning through experience. https://www.uipath.com/ai/agentic-ai This is an aspirational recommendation. No such Agentic AI exists in India’s DPI as of June 2025.

[13] Damaged credit bureau scores in rural areas represent a significant challenge in India. Defaults under SHG-bank linkage programs or JLG microfinance — often caused by circumstances beyond the borrower’s control or unforeseen contingencies — frequently lead to impaired credit histories. While comprehensive national statistics are lacking, the author’s field-level experience across multiple Indian states suggests that nearly one-third of applications from HWE entrepreneurs are rejected by banks and financial institutions due to past credit score damage.