HyperVerge Launches AI Agents to Speed Up MSME Loan Underwriting

HyperVerge Uses Three AI Agents to Turn Business Loan Underwriting From Hours Into Minutes

The company launched a new suite of three AI underwriting agents at Global Fintech Fest 2026 in Mumbai on September 9. The agents are designed to automate financial underwriting, assist with video-based personal discussions and perform background due diligence while keeping final lending decisions with human underwriters.

The approach is notable because HyperVerge is not positioning AI as an autonomous loan approver.

Instead, the system is designed to collect evidence, organize information, identify risks and direct important findings to human credit teams.

That makes the launch another example of AI moving from simple assistance toward specialized agents that perform complete sections of professional workflows.

What HyperVerge's Three AI Agents Do

HyperVerge has divided the underwriting workflow into three major areas.

Financial Underwriting Agent

The first agent focuses on financial documents.

It can work with information such as:

Bank statements
GST filings
Tax returns
Other financial documents

The system extracts relevant information, identifies missing information and helps reconstruct the financial picture of a business.

Instead of an underwriter manually searching through multiple documents, the AI can organize the important information for review.

This can be particularly useful for MSME lending because financial records are not always presented in a standardized format.

Video Personal Discussion Agent

The second agent focuses on borrower discussions.

HyperVerge says this system can conduct or assist with video-based personal discussions in 10–12 languages.

The agent can ask questions based on risks identified during the financial review and record responses with timestamps.

This creates another source of structured information for the lender.

Instead of relying entirely on manually written notes from a borrower conversation, credit teams can have a recorded and traceable record of relevant responses.

Background Due-Diligence Agent

The third agent handles background checks.

It can screen businesses and their promoters against sources such as:

Corporate filings
Litigation records
Sanctions lists
Other relevant public records

Potentially suspicious findings can then be routed to credit teams for further investigation.

This is important because background verification can involve searching across multiple information sources.

An AI agent can perform much of that initial research while leaving decisions about serious findings to human professionals.

Why MSME Lending Is a Good Use Case for AI

MSMEs often need financing quickly.

A small retailer may need working capital to purchase inventory. A manufacturer may need financing to fulfill a new order. A growing business may need funds to expand.

But lenders also have to manage risk.

That creates a difficult balance.

If underwriting takes too long, borrowers may struggle to access capital when they need it.

If lenders automate too aggressively, they risk making decisions based on incomplete or incorrect information.

HyperVerge's approach attempts to address both sides by automating the information-heavy parts while keeping humans responsible for final credit decisions.

HyperVerge Says AI Can Reduce Hours of Manual Work

According to HyperVerge CEO and co-founder Kedar Kulkarni, an underwriter's note that could take around two hours to prepare can potentially be generated by AI in about a minute and then reviewed by an underwriter in roughly five minutes.

The company also says its AI is designed to direct underwriters toward important information rather than create new facts.

That distinction matters in financial services.

A lending system cannot simply produce a convincing summary.

The information needs to be traceable to the underlying evidence.

Traceability Is a Major Part of the System

HyperVerge says its AI-generated recommendations can be traced back to their underlying sources.

Those sources can include financial documents, video timestamps and public records.

The company also says it uses output guardrails, links flagged issues to the underlying document or record and maintains an audit trail for regulatory review.

This creates a more explainable workflow.

An underwriter can potentially see not only what the AI identified but also where the information came from.

Human Underwriters Still Make the Final Decision

One of the most important aspects of HyperVerge's launch is what the AI agents do not do.

They are not designed to independently approve or reject complex business loans.

Kulkarni told The Economic Times that human involvement remains necessary because business loans can involve many different data points and complex risk considerations.

That makes the system a human-in-the-loop model.

The AI handles information gathering and analysis.

The human underwriter remains responsible for the final judgment.

Why Human Oversight Matters

Credit decisions can have significant consequences.

A false positive could cause a lender to approve a risky borrower.

A false negative could prevent a legitimate business from receiving needed capital.

There can also be information that is difficult for an automated system to interpret without context.

Keeping a human in the process allows lenders to investigate unusual cases and override AI recommendations when necessary.

HyperVerge Fits AI Into Existing Lending Infrastructure

Another interesting part of the system is that HyperVerge is not asking lenders to completely replace their existing technology.

The company says the system sits on top of existing infrastructure.

It is also modular, allowing lenders to start with individual tasks such as document review, background checks or video intelligence before expanding the deployment.

That could make adoption easier.

Financial institutions do not necessarily have to transform their entire underwriting operation at once.

They can begin with one repetitive workflow and evaluate the results.

Modular AI Could Be Easier for Banks and NBFCs

This approach is becoming increasingly common in enterprise AI.

Instead of deploying one enormous autonomous system, organizations can introduce specialized AI agents for individual business processes.

HyperVerge's three-agent model follows that pattern.

One agent handles financial information.

Another handles borrower conversations.

Another handles due diligence.

Together, they form an AI-assisted underwriting workflow.

The Potential Impact on India's MSME Credit Market

HyperVerge says India's MSME sector faces a significant formal-credit gap.

The company cites an estimate that only around 41% of India's 8.7 crore registered MSMEs have accessed formal credit, while putting the MSME credit gap at approximately ₹25 lakh crore.

If underwriting becomes less expensive and faster, smaller loans could potentially become more economical for lenders to process.

This is one of the most interesting implications of the technology.

AI does not necessarily need to replace a credit manager to change lending.

It could instead reduce the cost of preparing each application.

Lower Underwriting Costs Could Expand Access

Traditional underwriting requires employees to spend time collecting documents, analyzing financial information and checking records.

That cost becomes harder to justify for relatively small loans.

If AI can automate much of the preparation work, lenders may be able to process smaller applications more efficiently.

That could potentially make more businesses economically viable for formal lenders.

However, this remains a potential benefit rather than a guarantee.

The actual impact will depend on accuracy, lender adoption, regulatory requirements and borrower outcomes.

HyperVerge Already Uses AI Across Financial Workflows

The new agent-based system builds on HyperVerge's existing AI infrastructure.

The company's current platform includes financial tools for bank statement analysis, background checks, business verification and underwriting. Its small-business lending platform says AI can analyze financial information and support underwriting workflows while allowing lenders to decide how much automation they want.

HyperVerge also offers AI-powered verification and underwriting products for financial institutions in India.

Its platform includes services covering identity verification, business verification, fraud prevention and credit underwriting.

The new agents therefore represent an expansion of an existing financial AI stack rather than an entirely separate product direction.

HyperVerge Is Not the Only Fintech Company Moving Toward AI Agents

The launch comes during a broader shift toward agentic AI in financial services.

TheInfoBytes recently covered BharatPe's agentic AI assistant for merchants, which can interact with more than 60 live systems.

BharatPe's Agentic AI for Merchants

That system focuses on merchant operations rather than loan underwriting.

The difference is important.

BharatPe is using agentic AI to help merchants interact with financial and business services.

HyperVerge is using specialized agents inside the credit assessment process itself.

Another recent development covered by TheInfoBytes is Intellect's MSOCK platform, which focuses on giving AI systems deeper enterprise context for banking software.

Intellect MSOCK and AI Enterprise Context

Together, these developments show how financial institutions are experimenting with AI at multiple layers of their operations.

What Makes HyperVerge's Approach Different?

The important part of HyperVerge's system is not simply that it uses AI.

Many financial institutions already use machine learning and automation.

The newer development is the move toward specialized agents that can perform multi-step workflows.

For example, a financial underwriting agent can collect information from different documents, identify missing information and prepare a structured view for the credit professional.

A video agent can conduct a discussion and record relevant responses.

A due-diligence agent can search multiple sources and identify findings that require additional attention.

These are closer to task-oriented AI agents than traditional single-purpose prediction models.

AI Agents Could Change the Underwriter's Role

If systems like this become widely adopted, the role of a credit underwriter could change.

Instead of spending much of the day searching through documents, an underwriter could receive a structured summary of the application.

The human could then focus more heavily on:

Risk interpretation
Exceptions
Complex financial situations
Business context
Final credit decisions

In other words, AI could potentially move underwriters away from administrative work and toward higher-value judgment.

That is similar to what agentic AI is attempting to do across other professional industries.

The Biggest Challenge Will Be Trust

Despite the potential benefits, financial AI systems need a high level of reliability.

A mistake in a marketing chatbot is inconvenient.

A mistake in a lending system can have financial consequences.

That is why traceability, audit trails and human oversight are particularly important.

HyperVerge's emphasis on connecting AI findings to source documents and records is therefore more than a technical feature.

It is a requirement for making AI useful in regulated financial environments.

Speed Cannot Come at the Expense of Accuracy

The goal of AI underwriting should not simply be to approve loans faster.

It should be to help lenders process applications faster without weakening risk controls.

That is why HyperVerge's human-in-the-loop model is significant.

The AI can prepare the evidence.

The underwriter remains accountable for the decision.

What Happens Next?

HyperVerge says lenders typically begin with smaller groups of underwriters who compare AI-assisted results with their existing processes before expanding deployment. The company also says lenders monitor repayment behavior for borrowers handled through the workflow.

That kind of gradual rollout could become important for enterprise AI.

Rather than trusting an agent immediately with high-impact decisions, organizations can measure its performance and expand its responsibilities over time.

The same principle is increasingly being used across AI coding, customer service, cybersecurity and financial services.

Final Takeaway

HyperVerge has launched three AI agents designed to accelerate MSME business-loan underwriting.

The first analyzes financial documents, the second supports multilingual video-based personal discussions, and the third performs background due diligence across business and promoter records.

The system is designed to reduce repetitive manual work while keeping human underwriters responsible for final lending decisions.

That balance could be important as financial institutions adopt increasingly autonomous AI systems.

The bigger story is not simply that AI can read bank statements or conduct borrower interviews.

It is that specialized AI agents are beginning to take responsibility for entire sections of professional workflows.

For MSME lending, that could mean faster underwriting, lower operational costs and potentially broader access to formal credit — provided the technology can maintain the accuracy, transparency and human oversight that financial decisions require.

FAQs

What did HyperVerge launch?

HyperVerge launched three AI agents for business-loan underwriting: a financial underwriting agent, a video personal discussion agent and a background due-diligence agent.