Created in August 2024, Emergent labs has established itself as one of the most advanced startups in the automation of software production. Mukund Jha and Madhav Jha are developing a platform where specialized AI agents design, plan, write, test and deploy complete applications using natural language instructions. A value proposition that goes beyond simple no-code and questions the way in which organizations structure their engineering.
The company has just joined Google’s AI Futures fund, launched to support startups capable of exploiting the most advanced AI models. The fund combines capital, early access to Gemini models and direct technical support. “The partnership provides us with three critical resources: capital to accelerate our growth, early access to Google’s Gemini 3 model, and operational technical support from Google’s AI team. Together, these elements clearly distinguish the AI Futures Fund from traditional investments,” Mukund Jha, co-founder and CEO of Emergent, told FW.Media.
As for the fund, the objective is to support the most promising AI platforms: “Emergent allows companies to realize their ideas, remove obstacles and democratize access to the tools necessary for building their technological infrastructure,” underlines Jonathan Silber, director of Google’s AI Futures fund.
This alliance comes in a phase of hypergrowth for the startup. In less than five months, Emergent claims 2.5 million users and annual recurring revenue in excess of $25 million. “We already serve 2.5 million users globally and have reached $25 million ARR in less than five months. This partnership gives us the resources and technological base to scale to tens of millions of users by next year,” says Mukund Jha. A trajectory that did not fail to convince LightSpeed Ventures to invest $23 million, following a $7 million seed round with together fund and YCombinator
The platform is based on a multi-agent architecture that operates as a complete engineering team. Each agent explores the code, navigates the tree, identifies dependencies, runs tests, and fixes errors. To guarantee this autonomy, Emergent has built its own technical stack: execution environment, deployment pipeline, database, security, Kubernetes orchestration, an important difference compared to no-code platforms which are limited to visual composition.
This approach reconfigures the software cycle with fewer repetitive tasks, an acceleration of time-to-market, increased productivity and a direct impact on the structuring of technical teams. Access to Gemini 3 opens up an additional field of experimentation for Emergent. “Access to Gemini 3 is decisive for our product development. It allows us to explore the limits of agentic AI and increase the sophistication of the applications our users can build. With direct support from Google AI experts, we can integrate these capabilities faster,” observes Mukund Jha.
The platform has already made it possible to create various applications such as CV management tools in the United Kingdom, a marketing audit application in Germany or even a directory of AI use cases to help with the operational adoption of emerging technologies. “The partnership with the AI Futures Fund aligns perfectly with our mission: to democratize software creation so that everyone can transform an idea into a functional application, without prior technical skills,” concludes its founder.
However, several weaknesses remain to be monitored. First, the sustainability of an ARR built very quickly on a large base of “tourist” users which will depend on the ability to sustainably grow the share of revenue from power users actually engaged. Then, behind the discourse on the “world-class agent”, Emergent remains structurally dependent on third-party models for code generation, with cost, performance and licensing issues that the in-house infrastructure does not resolve. The company is also operating in an environment that is already very crowded with solutions like Replit, V0, and historic no-code platforms, which requires it to maintain a tangible product differential, both in terms of the quality of the apps produced and the mechanics of distribution via influencers. Finally, the choice to operate its own infrastructure layer (VM, back-end, bases) offers fine control and better feedback loops, but at the cost of high operational complexity in terms of reliability, security and scalability, which will remain an ongoing project at this level of growth.
