Building Where Others Don’t: How Hardik Bansal Turned Neglected Markets Into Scale

The pattern in Hardik Bansal's career is easy to see in retrospect: he picks problems the technology industry has collectively decided are too hard or too niche to be worth solving, and then builds a company with precise, empathetic technical infrastructure around them.

In 2019, that meant starting a dating-and-social company for India's Tier-2 and Tier-3 cities, a market where every major investor had already written off the category. Today, it means building an AI knowledge layer for municipal government data, an asset class that institutional investors widely acknowledge is important but that remains almost entirely unstructured and inaccessible at scale.

The connecting thread, in Bansal’s telling, is personal. “Most consumer products are designed for the majority of spend, in the wealthiest cities, in the most familiar formats,” he said. “I have always been more interested in the gaps. Where is real value being left on the table because the right system for it has yet to be built?”

Bansal grew up in India and trained as a computer scientist. His engagement with generative machine learning started early: in 2017, years before the foundation-model wave, he published an implementation and an explanatory guide for CycleGAN, an unsupervised image-to-image translation framework that has since garnered academic citations and become a reference in the field. After early work spanning theoretical computer science and finance, he turned his attention to a problem that was obvious to anyone who had grown up outside India's major metros.

The country's mainstream consumer apps were designed for a narrow slice of urban users. Dating apps were the starkest example. Imported swipe-based interfaces assumed a cultural openness to casual online interaction that Tier-2 and Tier-3 cities had never adopted. Women faced safety concerns and social stigma that the swipe-based model was structurally incapable of addressing. Investors had repeatedly lost money trying to force Western templates into the Indian market.

In 2019, Bansal co-founded FRND with a design thesis that ran against every prevailing assumption in the category: replace the photo-based swipe with real-time, moderator-guided audio interaction. Users would be matched into voice rooms with active moderation. Conversation could happen safely. Anonymity would be preserved until both parties chose to move forward. There were no profile photos at all.

"It was orthogonal to what every other dating app was doing," Bansal said. "We had moderators. We had voice. From the outside, it looked like a worse product. From the inside, we knew it was the only product that could actually work for the user we cared about."

The underlying system, a real-time moderated audio interaction mechanism for dating with no precedent in the global social app landscape, was patented as a novel approach. The innovation had a measurable social impact: it gave millions of women in Tier-2 and Tier-3 Indian cities their first safe entry point into online social interaction, addressing safety and stigma barriers that every previous platform had left intact.

Making it work at scale required Bansal's team to build one of the earliest production deployments of generative AI in consumer moderation. The pipeline processed enormous volumes of unstructured speech in real time across 10 different regional languages, identifying safety risks, interpreting cultural-context cues, and adjudicating edge cases at a speed and volume that human review teams could never have matched. The team was shipping production generative AI years before the technology entered mainstream enterprise conversations.

The bet paid off. Within three years of launch, FRND scaled to more than 50 million users and crossed $25 million in annual recurring revenue, attracting an $8.5 million investment from Krafton, the South Korean publisher behind PUBG, and Elevation Capital. In 2021, Bansal was named to the Forbes 30 Under 30 Asia list in the Consumer Technology category for his role as co-founder and chief technology officer.

From B2C in Bharat to B2B in the U.S.

By the time FRND had established itself as a category leader, Bansal had begun thinking about what to build next. The answer came from a different continent.

He joined Turl Street Group LLC as Chief Technology Officer of GatherGov, the company's flagship product. Where FRND had addressed a gap in how Indian consumers connected with one another, GatherGov addresses a gap in how institutions access the public data that shapes their most consequential decisions. The platform converts local government meetings across thousands of jurisdictions into structured, queryable intelligence for real estate developers, institutional investors, and major newsrooms. It is venture-backed by ComposeVC and OPCO Ventures.

The transition from B2C consumer tech in India to B2B enterprise AI in the United States was, by Bansal's own account, among the hardest passages of his career.

"You need a completely different skill set," he said. "The motion of selling to a Fortune 500 institutional investor is nothing like the motion of acquiring a 25-year-old in a Tier-2 Indian city. The problem statement is different, the customer is different, and the solution is different. I had to relearn a lot of what I thought I knew about how products meet markets."

What carried over was the conviction that mattered most: build for the user the industry has overlooked, design for scale from day one, and treat every system as permanent infrastructure rather than a proof of concept.

"At FRND, the overlooked user was a 22-year-old woman in a mid-sized Indian city who had never felt safe on a dating app," Bansal said. "At GatherGov, it is an analyst at a major institution who knows that critical information is being discussed in thousands of municipal meetings every week and has no way to access it systematically. The specifics changed completely. The question I ask myself stays the same: who is being underserved, and is the technology now ready to do something about it?"

The quote that best captures Bansal's decision-making framework is deceptively simple. "I have always been drawn to problems where the technical opportunity and the human opportunity arrive at the same time," he said.

At FRND, the convergence was mobile penetration in India reaching a level that made a 50-million-user social product viable, at the same moment that moderation-grade machine learning became affordable enough to deploy at scale. At GatherGov, it is the post-pandemic mandate that pushed virtually every U.S. municipality to record and publish meetings, arriving alongside foundation models and graph-based retrieval capable of extracting structured intelligence from that archive.

In both cases, the opportunity had technically existed for years. What changed was the economics. And Bansal argues that founders who wait for a category to be validated before entering it will always be building the same product as everyone else, competing on distribution rather than on insight into the user.

The Organizational Discipline That Compounds

When asked about his most important lesson from FRND, Bansal points not to technology, but to leadership.

"Early on, I wanted to do everything myself," he said. "I was aiming for perfection on every task, and that made it genuinely hard to hand over problem statements to other people. In the early days, you can get away with it. Over a sustained period, it breaks."

The shift required redirecting his impulse for perfection away from individual output and toward the team itself. Build a team that can execute at the standard you hold yourself to, and then focus your energy on making that team better rather than doing their work.

"That was the most important transition of my career," he said. "The job of a CTO at scale is to build the team that writes the best code in the building. If you are still trying to be that person yourself, you have already become the bottleneck."

That philosophy now shapes how he is building GatherGov's engineering organization. The bar is set by the company's institutional users, who expect accuracy, consistency, and citation-grade traceability. Bansal's focus is on assembling a team that can meet that standard independently, across every layer of the platform, without depending on any single engineer's heroics.

His longer-term ambition extends beyond any single product. "I see myself building and leading a team that can operate across both engineering and business, with a shared focus on speed, clarity, and impact," he said. "The technology problems will keep changing. The organizational discipline is what compounds."

Bansal's trajectory is unusual enough that observers frequently ask him to define his overarching strategy. He resists the framing.

"If there is a strategy, it is a single question," he said. "Who is being underserved, and is the technology now ready to do something about it? Every time I have answered that question honestly, the next company has appeared. I expect that to keep being true."