Krutrim's Transformation to Profitability

550 employees → 150. Chatbot pulled. Chips paused. And they just posted their first profit. This is Krutrim's story. In 2024, Bhavish Aggarwal launched India's first AI unicorn with one mission — build India's OpenAI. Multilingual LLMs. Proprietary chips. An agentic assistant called Kruti. Consumer AI products. The whole stack. Investors loved it. $50M raised. $1B valuation. India's fastest unicorn ever. Then reality hit. Building competitive LLMs needs world-class researchers and massive GPU farms. Chip design needs billions and decades. Consumer AI needs trust — which erodes fast when your chatbot gives wrong answers. By late 2025, Krutrim had tried to fight OpenAI, Google, and TSMC at the same time. Something had to give. So they made the hardest call a founder can make: Kill the vision. Save the company. Chip design — paused. Foundation models — paused. Kruti — quietly pulled from app stores, no announcement. They went all-in on one thing: AI cloud infrastructure. Built in India. For Indian enterprises. The result? ₹300 crore revenue in FY26. 3x growth year-on-year. First ever annual net profit. Margins over 10%. Self-funded. No new round needed. Most founders would rather burn the company defending a big vision than admit the market needs something smaller. Bhavish didn't build India's OpenAI. He built something rarer — a profitable AI company. Have you ever had to kill a product to save the company? What did that decision look like? #StartupIndia #AIStartups #Entrepreneurship #Krutrim #Founders