On August 15, 1999, Srikanth Velamakanni made an unexpected decision over a meal with a friend: he would quit his job by January 1, 2000, and start a venture. The call surprised even him. At the time, he had no clear business plan and little inclination towards entrepreneurship. “I am going to quit…and start a venture,” he recalled saying, almost impulsively.
The hesitation ran deeper. Growing up in the 1980s, he had internalised the idea that business and ethics rarely coexisted. That assumption began to shift during his time at Indian Institute of Management Ahmedabad, where a guest lecture by N R Narayana Murthy challenged that belief. Murthy’s account of building Infosys on ethical foundations altered Velamakanni’s view of entrepreneurship. Around the same time, peers raising venture capital demonstrated that personal wealth was not a prerequisite to start a company.
The idea that would eventually define Fractal, however, had older roots. Velamakanni studied engineering at Indian Institute of Technology Delhi, drawn by a strong interest in mathematics and probability. At the same time, he was equally interested in behavioural sciences, taking courses on the psychology of human decision-making. The convergence of these two interests — quantitative models and human behaviour — would later become central to the company’s approach.
That convergence became clearer during his stint at ICICI Bank. He observed how global credit scoring models were being adapted for India with manual adjustments. Velamakanni believed statistically validated, locally calibrated models could do better. The broader insight was that data and behavioural understanding could be combined to improve decisions at scale.
Within months of his decision, he brought together a group of co-founders, including Pranay Agrawal and former classmates. Each committed Rs 2 lakh as seed capital. For some, including Velamakanni, this meant borrowing from family. “My mother lent me the seed capital,” he said. In early 2000, they quit their jobs and began operations, working long hours with minimal resources.
Fractal was formally launched in February 2000 as a consumer Internet venture aimed at improving buying decisions through a comparison shopping platform. The timing proved difficult. Within months, the dotcom crash made the model unviable. By November 2000, the founders concluded the business would not survive in its original form.
Dotcom Pivot
The pivot that followed defined the company. Fractal shifted to analytics — using mathematical models and behavioural insights to help enterprises make decisions. What began with credit risk modelling expanded into marketing analytics, consumer behaviour, and eventually artificial intelligence. “We started by powering consumer decisions, and today we power enterprise decisions,” Velamakanni said.
Early funding was limited and uncertain. The founders turned down an initial investment offer due to unfavourable terms, shortly before the market downturn made capital scarce. Support from friends and early backers sustained operations until 2001, when an angel round led by industry investors provided runway to stabilise and validate the analytics model. Institutional capital came much later, in 2013, when TA Associates invested $25 million.
Over time, Fractal evolved into a global enterprise AI and analytics company, serving clients across sectors with data-driven decision systems. It now operates across multiple geographies, including the US, the UK and India, with headquarters in Mumbai. The company has raised about $685 million from investors including Apax Partners, TPG Capital Asia and Khazanah Nasional. In FY25, it reported operating revenue of Rs 2,765 crore.
Scaling Through Crisis
The trajectory has not been linear. Between 2006 and 2008, the company faced a leadership crisis following differences over strategy after appointing an external CEO. The period saw team exits, client uncertainty and legal disputes. At one point, about 30 employees left in a week from a base of around 80. Velamakanni said the episode forced the company to rebuild and impose greater discipline.
Subsequent milestones marked its recovery and expansion. Early work with HDFC Bank in 2001 helped establish credibility in risk analytics. In 2012, Fractal set up a dedicated R&D unit to deepen its capabilities in data science and AI. More recently, the company went public, listing on the BSE and the NSE in February 2026.
For Velamakanni, the original idea remains intact despite changes in form. The company’s focus continues to be on improving decisions, first for consumers and now for enterprises. The method, which is combining mathematical models with an understanding of human behaviour, has scaled from early credit models to full-spectrum AI systems.
Looking back, the venture that began without a clear plan has evolved into a data and AI firm with global operations. The shift from a failed Internet startup to an enterprise analytics company was less a change in direction than a refinement of the underlying idea.