The Brothers Who Grew Up With Code
Like many kids of that time, their initial excitement came from video games. But while others played to win, the Jha brothers wanted to understand how the game worked. They were curious about what lay beneath the screen and what logic made the pixels move. This pure, almost innocent curiosity became the foundation of their lifelong connection with software.
Long before startups, funding rounds, or discussions about AI, there were just two brothers sharing ideas, breaking things, fixing them, and diving deeper into the world of code.
Two Paths, One Obsession
After their post-graduation in India, both brothers left for the United States to pursue PhDs in computer science at different universities. Their journey was different from the very beginning. Madhav completed his PhD, while Mukund made a difficult and risky decision. He dropped out after one semester.
In 2010, Mukund started working at Google as an intern.
This was the turning point in his career, which was followed by many such turning points. Over the years, the brothers would become employees of the world’s most powerful technology companies—Google, Amazon, Dropbox—thus gaining firsthand exposure to the building, testing, and deploying of large-scale software systems. Mukund, however, would come back to India and start Dunzo, one of the most popular hyperlocal startups in the country.
The years passed by with their lives being full of work and other commitments. However, one thing that never went away was the fact that even when their careers took different paths, Mukund and Madhav kept a strong bond through their common intellectual curiosity. They used to share their notes, argue their points, exchange snippets of code, and follow the journey of machine learning together and most of the time it was very late at night after work and life had their rest.
When the Code Began Writing Itself
In 2018, something changed.
First the brothers were differently uniquely excited by the technology development, then they followed OpenAI's early large Language models, which in function opened a brand new horizon of what they expected from software. The new software was not only faster or able to give better tools, but it was software with reasoning, generation, and more and more writing code capabilities.
During the following five years, the number of AI-powered platforms increased at an incredible rate. While a great part of the world was still in a dilemma about whether AI would replace programmers, the Jha brothers were only thinking about a completely different question: What if software is the one that becomes accessible to everyone?
Mukund has the memory very strong. Their continual immersion in AI amidst their personal and professional lives was a growth pattern rather than a sudden realization. "We were super nerds about this," he pointed out. "We'd sacrifice our sleep hours just to learn more. At one time I phoned Madhav and said, 'Why don't we create something together?'"
The founding of Emergent Labs was thus heralded by that phone call.
From Collapse to Convergence
Dunzo was going through a tough time in 2023 as well as 2024 and eventually decided to shut down permanently. Mukund, on the other hand, was going through a change and was having some deep personal reflections as well. When one chapter was closed with the shut down of the company, another chapter with the idea of software changing the world in the most radical way was opened.
The two brothers came up with one single and brave idea by the start of 2024: employing AI for coding automation and software testing automation. They went to Y Combinator with this idea and without a business plan in detail. They had life experience, technical skills, and trust instead of a business plan.
The brothers are behind Emergent Labs, a startup that is now part of Y Combinator Summer 2024 cohort. The startup was able to raise a $5 million round from Girish Mathrubootham’s Together Fund in a short period of time. In addition, to the cash, the support gave them another thing which is equally important - clarity.
Instead of doing the hard work of automating software testing completely, the brothers took a step back and asked a simpler question - what is the smallest meaningful unit of software development that AI can automate today? The answer was not testing. It was building apps.
Proving That Automation Wasn’t a Fantasy
Emergent was originally an AI agent whose main goal was to locate and fix bugs in the code. To see how far they could take it, the team measured its performance against the newly introduced Software Engineering Benchmark (SWE), a standard that challenges AI agents to independently solve 500 real GitHub issues.
In just two months, Emergent was at the very top of the SWE leaderboard.
For the founders, it was not about the rankings. It was the confirmation. Seeing their AI agents to be capable of independently solving complex engineering problems made one thing obvious: software automation is no longer a thing of the future—it is already happening.
“It showed us the direction where the world was going,” Mukund, CEO of Emergent, at present, recalled. “This was the only logical step after this one.”
Building for Those Who Don’t Code
A private beta for Emergent was made available in February 2025 and a public launch followed in June. The platform had managed to raise a total of $30 million and attract 3 million users by the middle of 2025. However, Mukund refers to this period as "barely scratching the surface".
The mission of Emergent is very different from that of platforms like Claude Code, OpenAI’s Codex, Gemini AntiGravity, or Cursor. The brothers were not figuring out how to make a better developer tool. They were figuring out how to eliminate the need for developers altogether.
Emergent is a platform that was built from the ground up for people without a technical background. There are no compiler errors that need to be understood, no logs that need to be manually debugged. The AI agents are taking care of these issues silently in the background, thus the users are allowed to concentrate on what they want to create rather than how to create it.
The point was straightforward, but it was a revolutionary idea as far as its consequences were concerned: software engineers ought not to be the only ones who have the right to develop apps.
Growth, Users, and a New Kind of Builder
Within just eight months after its beta launch, Emergent expanded to a team of 40 employees who are located in both India and the United States. Out of its 3 million active users, roughly 60,000 are paying customers.
The majority of the users are solo founders, small business owners, and first-time entrepreneurs. They turn to Emergent to create logistics tools, marketplaces, chatbots—software that would have taken them months of development or a hefty engineering team to make.
Emergent is subscription-based, with tiers that cost between $20 and $200 a month depending on the user's consumption. Besides that, users pay for hosting to run their apps on Emergent's platform. Given the current momentum, the startup is projecting an annual recurring revenue of about $25 million in FY26.
The team is also looking into other revenue streams such as improved app discovery for apps built on the platform—just like products are surfaced by marketplaces today.
The Frictions That Still Remain
Not everything is perfect or easy so far.
Deployment is one of the major problems for Emergent—and other “vibe coding” type platforms”—as to where the technology goes. It has been made easier to create apps, but publishing AI-generated apps on Google Play Store or Apple’s App Store is still a bit complicated. At the moment, only around 15-16% of Emergent users have their mobile apps live.
Mukund admits the difference. The team is working hard to make this layer simple, they understand that real accessibility does not stop at creation, it has to go further to launch and scale.
To start with, Emergent is committed to strengthening its infrastructure, enhancing data security, and keeping up with the rapid changes of AI models.
Competing With the Giants
The competition is getting heated as AI is changing each layer of software development. The global AI discussion is mostly about such platforms as Anthropic’s Claude, OpenAI’s Codex, and Google’s Gemini, which are supported by deep capital and huge ecosystems.
Mukund sees the risks very clearly. As AI models get better in a world where they exist, the distinctions between the players become thinner. How long to be relevant will depend on how fast the platforms can innovate, deliver value, and stay close to real user needs.
Emergent Labs doesn’t have a long history yet. Its goal—app development should be easy for anyone—was very clear. The question that is left is not how AI will change software but whether smaller, focused teams like Emergent can be as fast as the giants who are shaping the future of AI.
At present, the brothers are following their old pattern - they are still figuring out what the future might be and constructing it.