How SwiffyLabs’s Modular Lending Platform Enables Fast-Scaling Fintechs

At ZestMoney, the goal was to democratize credit for India. Within 24 months of our hypergrowth phase, we were disbursing over $2.6M every month. Twelve months later, that was $4M+. The loan book was scaling fast. The tech underneath it, wasn't though. Our lending stack, from loan origination to post-disbursal monitoring, couldn't keep pace. Outages were routine. I was watching the team burn weeks firefighting systems instead of building, and the realization landed hard, the technology meant to power our growth had become the thing limiting it. So I went looking for a replacement. Every vendor I evaluated pulled me back to the same wall, legacy LOS and LMS platforms built for the loan books of a decade ago. Rigid systems where the smallest policy change took weeks, with no clean way to plug into the digital ecosystem a modern lender runs on - bureaus, KYC providers, payment rails, collections partners. They couldn't keep pace with scale, and they couldn't move fast. Switching meant ripping out and rebuilding everything. Months of work, resources I didn't have. Then it clicked, I wasn't alone. The 10,000 formal lenders in India across banks, NBFCs, and fintechs were all running the same legacy LOS and LMS systems, all hitting the same ceiling. I'm a techie at heart, so I couldn't leave it alone. I set out to build a modular lending platform that drops into a lender's existing stack instead of forcing a rebuild. KYC breaking most often? Add the KYC module. Underwriting accuracy and routing weak? Use ours. And for a lender starting fresh or ready to replace the core, the same modules come together as a full stack, from origination to collections. That became SwiffyLabs.

What worked, talking. Not pitching, listening. My team and I spent our early months in conversation with the people who actually live inside an LOS and LMS every day, credit ops, risk, collections. What slows them down. Where the system fights them. What a clean workflow would actually look like. We'd run lending operations ourselves, so these weren't survey calls, they were peer conversations. Two things came out of it, the product got sharper because real practitioners shaped it, and pipeline followed because a lender who feels understood takes the next call. What flopped, trying to shortcut that with scale too early. We ran outbound against a purchased list, generic messages to names we had no context with. It was the opposite of what had been working. The same kind of lender who engaged in a real conversation ignored a cold, templated send. It didn't sink us, but it taught the lesson cleanly, in lending, context comes before the pitch. Earn the conversation, then have it.

The most common feedback isn't about the platform itself. It's about change. When a credit ops or collections team has run the same screens for ten years, moving to anything new takes adjustment, even when the new system is faster. The muscle memory is real. A field they expect on the bottom right now sits top right. An export format looks slightly different. None of it slows the loan book, but it's the friction a user feels on day one. We treat that as our job to absorb, not theirs. Every customer gets a dedicated resource who maps the rollout to how their team already works. Where it helps adoption, we configure the platform to mirror their existing workflows rather than force a new one. That comes with detailed documentation and real knowledge transfer, not a link to a help center. Support runs 24x7, and resolution is fast. The technical go-live stays quick. The human side of the switch gets the hand-holding it actually needs, and adoption follows because the team feels supported through the change rather than thrown at it.

The biggest misconception is that legacy lenders are content with the status quo. The story everyone tells is that banks and large NBFCs have made their peace with old systems, that speed is a fintech obsession, and the incumbents don't really care. That's wrong, and the numbers say so. Look at SBI. The largest bank in the country, the definition of an incumbent, grew its gross advances by roughly 17% this year to about $50B+. That is not the behavior of an institution that thinks it has stopped competing. A loan book that size growing at that pace is hungry, and it's hungry in exactly the places where technology decides who wins, underwriting speed, disbursement TAT, collections efficiency. The funded fintech LSPs get the attention for moving fast, but the banks and NBFCs are chasing the same scale, and they know legacy LOS and LMS are what hold them back. Some of my best conversations are with business heads inside these banks who already understand that the constraint on their growth is technology, not appetite. The people who get this market wrong are still picturing a bank that no longer exists.

The next chapter is SwiffyLabs Studios, our AI-native product studio for banks, NBFCs, and fintech lenders. The lending stack solved the core systems problem. Studios goes a layer up, helping institutions design, build, and ship financial products at the speed their loan books demand. Over the next 6 to 12 months, we're putting AI into the build itself. Agentic engineering that turns requirements into tested, secure, compliant code. An application builder so teams ship app changes without re-shipping the app. Voice-led KYC and loan origination. Chat-to-action onboarding across WhatsApp and web. All of it built compliance-first for RBI and DPDP because in lending, speed without controls is a liability. The bet is simple. The lenders who win the next decade will out-build everyone else, and we want to be the reason they can.

In our first 12 months, SwiffyLabs has processed over 9M loan journeys and now powers a loan book of more than $1B+. Most new entrants take three to five years to reach that scale. We're targeting a 10x loan book over the next two years.

I wasn't starting from scratch, and I wasn't guessing at the problem from the outside either. I've spent more than two decades in financial technology, long enough to watch the industry move from branch ledgers to real-time digital lending. The part that mattered most came at ZestMoney, the lending business I founded. Running a lending business teaches you what nothing else can, what breaks at 2 a.m., what a rising NPA does to a quarter, how much disbursement TAT actually costs you in lost customers. So when I started SwiffyLabs, I'd already lived on the other side of the table, as the buyer frustrated with my own loan stack. The team around me comes from the same world - people who built core banking and lending platforms at places like Finacle and ZestMoney. That combination, builder and operator, is why we knew exactly what to fix first.

Early on, I assumed a faster, more reliable system would sell its own adoption. Build something clearly better than the legacy stack, and teams would move to it. That was wrong. What I underestimated was the human side of the switch. A credit ops team that's run the same screens for a decade doesn't adopt a new system because it's technically superior. They adopt it when the change is made easy for them. Our early rollouts moved faster on the engineering than on the people, and that gap showed up as slow adoption even where the platform was working perfectly. The lesson reshaped how we deliver. We now treat onboarding, configuration to existing workflows, and dedicated support as part of the product, not an afterthought. Go-live gets a customer started. Adoption is what makes the technology pay off.