Shriram Finance sharpens in-house tech focus, pushing build-to-buy ratio to 80:20

In 2022, Shriram City Union Finance and Shriram Capital merged with Shriram Transport Finance, forming the non-banking financial company Shriram Finance Limited. Along with getting the business process in order, the Mumbai-headquartered organisation embarked on a digital transformation journey, building on its half-century-long legacy.

From the beginning of this digital overhaul, the company was careful not to fall trap of using digital transformation as another industry buzzword, Vinod Kumar, the chief digital officer at Shriram Finance, told TechCircle. “We were very clear about our intent. We knew we were a 50-year-old company with deep roots and strong foundations, but we also wanted to be future-ready and truly digital-first for the next 50 years.”

To be sure, Shriram Finance is primarily focused on retail asset financing, serving small road transport operators, small business owners by offering a range of credit solutions.

Physical systems to digitised workflows

Traditionally, the company has operated through a heavily physical model, both in terms of customer outreach and engagement. It had a large on-ground workforce that was physically available at every customer touchpoint: from onboarding, document collection and validation, to disbursal and post-loan servicing. But now, the entire physical journey has evolved into a digital workflow. 

“Take our recent tool, for example, it enables end-to-end digitisation of the customer journey. Once a customer reaches out to us (often through our B2C assets), our teams can instantly engage. There’s no longer any need to physically collect documents. Everything can be gathered digitally through integrations like eKYC. Validations, which earlier required physical checks, are now performed online,” said Kumar. 

Further, in the older model, the files had to be moved from the dealer location to the branch office, and the underwriter had to manually inspect physical paperwork. This has been eliminated, and all of this now happens within the system. 

Even the loan agreement process, which used to be laborious and time-consuming, is now digital. A customer no longer needs to physically sign a 36-page document. Instead, they receive a digital link, review the terms, and complete the process using just a couple of OTPs. The signed agreement is automatically stored and can be retrieved at any time. What once took 24–48 hours to process, especially in the two-wheeler loan segment, can now be completed end-to-end in 20–25 minutes, explained Kumar.

“We’ve embedded smart calculators, credit bureau access, and validation tools directly into the platform. This allows our teams to fetch and verify data, run customer profiles through scoring models, and generate sanctions quickly,” said Kumar.

Shriram Finance has also announced a SuperApp, which Kumar claims is the only fifth or sixth NBFC to do so in India. It is now a key channel for onboarding, servicing, and cross-selling, significantly improving customer access and speed of service.

“Once we had B2C sorted, our focus shifted to internal transformation. We saw our own employees as the next critical customer segment. So we began revamping our internal platforms, starting with the LOS (Loan Origination System), the LMS (Loan Management System), and breaking down journeys by product lines,” said Kumar.

Strategic approach to AI

Like with digital transformation, Kumar said that his team did not want to get lost in the rigmarole of buzzwords and hype with AI. “For us, AI had to be purposeful, rooted in outcomes. And we knew from the start that meaningful AI is only as good as the data behind it.”

A large part of leveraging AI in its processing was spent on transforming raw data into insights. Sitting on a heap of data collected over the years meant that a lot of it was unstructured and siloed. A customer with multiple products might have appeared in different datasets with no linkage between them. And in some cases, the team had to deal with incomplete records.

“We invested a lot of time and effort in cleaning, enriching, and organising our data, because without that foundation, we would’ve had to rely on external or alternate data sources, which we wanted to avoid.” This set the groundwork for AI, followed by running multiple A/B pilots across the customer journey.

“One of our early wins came from a propensity model built using our own data and bureau insights. This model predicted when a customer was likely to take a loan, the probable quantum, and the product type. This allowed us to pre-score and pre-approve customers, and push personalised marketing offers at just the right moment,” noted Kumar. 

The team has now built a dynamic pricing engine that pulls data (over 100 parameters) from various sources/ This system determines whether the customer should be given loan and at what interest rate, basis their risk profile. The company also used an early warning signal (EWS) engine that continuously monitors customer repayment behavior and throws up anomalies, predicting possible defaults. 

“We’re also using AI in customer service. Given the size of our customer base, physical servicing was becoming increasingly difficult to scale. So we introduced AI-powered bots to respond to emails, auto-classify requests, detect sentiment, identify language, and respond intelligently. This has drastically improved both our speed and quality of service.”

Kumar said that the entire AI-related work is done in-house. 

Layered approach to cybersecurity

Shriram Finance’s cybersecurity framework built on three core pillars: asset protection, data security, and network resilience. For our customer-facing digital assets, including web and mobile platforms, the company has implemented advanced cyber attack surface monitoring tools that proactively scan for vulnerabilities. It is complemented by a live AI-driven engine designed to detect and block fraudulent users in real time.

The team implements data masking and encryption protocols across data centers, meeting regulatory requirements to prevent unauthorized data exposure. Additionally, a database activity monitoring system that functions as a firewall for our databases continuously tracks access to sensitive information and flags any anomalies for immediate action.

From a network security standpoint, the team has built a layered defense strategy that includes virtual firewalls, network segmentation, and a zero-trust architecture.

Building in-house vs outsourcing

Of late, the scales in the banking, financial services and insurance (BFSI) industry have tipped in favor of building IT and technology solutions in-house rather than outsourcing them. Shriram Finance, Kumar notes, is aligned with this shift.

“We prefer to develop most enhancements internally, especially when it comes to platform-level improvements. However, we’re also pragmatic; it ultimately comes down to value for money and the speed at which we want to go to market. That’s where we sometimes partner externally,” he said. The tech team comprises 350 staffers.

Typically, the only areas where the company looks outward are where integration with a central repository is required, for example, API-based validations. 

“If I had to quantify it, our build-to-buy ratio would be around 80-20. Every initiative starts with an in-house attempt, and we only explore external solutions when there’s a clear advantage in terms of speed, cost, or operational efficiency.” This approach allows the company to stay agile and reduce dependencies, while still being open to smart integrations.