Bhavesh Aggarwal’s Krutrim showcases how the AI ambitions of India are shifting from models to infrastructure.
Bhavesh Aggarwal-led AI startup named Krutrimis now eyeing cloud infrastructure services after focusing on building large language models and AI chips. This serves as a major move by the startup[internal link of website] and a change in strategy in less than two years of becoming India’s first AI unicorn, as per an Economic Times report.
The evolving strategy of Krutrim.
The country’s AI ecosystem has focused on creating indigenous AI models, developing sovereign computing capabilities, and reducing dependence on foreign technology platforms. However, as the economics of artificial intelligence become clearer, many companies are reassessing where the biggest opportunities truly lie. This changing reality is reflected in the evolving strategy of Krutrim, the AI startup founded by Bhavish Aggarwal.
Rather than concentrating only on building AI models, the company has now increasingly shifted its attention toward AI cloud infrastructure and enterprise computing services. The move points to a broader trend taking shape not only in India but across the global AI industry.
The reality of building AI.
Developing advanced AI systems requires far more than technical expertise. It demands enormous investments in graphics processing units (GPUs), data centers, cloud infrastructure, engineering talent, and ongoing research. While breakthrough AI models often attract public attention, maintaining and scaling them can be extraordinarily expensive. For many emerging AI companies, generating sustainable revenue from these investments remains a significant challenge. As a result, industry observers are increasingly questioning whether developing foundation models alone is enough to build a profitable long-term business. The focus is gradually shifting toward the infrastructure that powers those models.
AI can run on computing power. Every AI model, chatbot, enterprise application, or autonomous agent depends on vast amounts of infrastructure operating behind the scenes. As businesses across sectors such as healthcare, finance, mobility, retail, and telecommunications adopt AI solutions, demand for computing resources continues to grow rapidly. Many organizations thus prefer to access these capabilities through cloud platforms rather than invest heavily in their own infrastructure. This has created a growing market for AI cloud services, where companies provide the computing power needed to train, deploy, and operate AI systems at scale. For providers, infrastructure services often offer a more predictable and sustainable revenue model than the highly competitive race to build new foundation models.
Krutrim’s shift showcases how developments occur throughout the international AI landscape. All over the world, tech companies are investing heavily in AI infrastructure, specialized chips, cloud platforms, and computing networks. Increasingly, the conversation is moving beyond which company has the most advanced model to who can deliver AI efficiently, reliably, and at scale.The evolution of companies like Krutrim signals a more mature phase in India’s AI journey. Building homegrown AI models remains an important national objective, but industry leaders are increasingly recognizing that models alone do not create a thriving AI ecosystem. Cloud infrastructure, computing capacity, enterprise adoption, and sustainable business models are equally important pieces of the puzzle.