Fractal Analytics has two agentic AI projects in development, a first for the company, one of which has already started pulling in revenue, according to company co-founder Srikanth Velamakanni.
“We aim to deploy agentic AI systems to drive enterprise productivity, improve accuracy (and effectiveness) of decision-making and improve customer experience,” Velamakanni said in an interview with Mint.
Agentic AI systems are an emerging technology within artificial intelligence that combines the capabilities of large language models, machine learning and natural language processing to perform tasks independently and with minimum human supervision.
In 2024, enterprise applications with built-in agentic systems have only 1% adoption among companies in 2024, but that number is likely to go up to 33% by 2028, according to management consulting firm Gartner. Fractal’s shift towards agentic AI systems marks a shift in the company’s revenue generation strategy. “All of Fractal’s revenues are from building AI systems to power decisions. We expect many of these to become agentic in the near future,” Velamakanni said.
Delivering insights
Fractal’s deployed agentic AI project is called ‘Cogentiq’ and is an agentic AI platform that accesses a company’s data to deliver insights to assist in decision-making and execution. “We’re creating a marketplace of agents that work with each other or can be used in any problem-solving that we do for clients,” Velamakanni said.
The company expects that licensing the Cogentiq platform will contribute 20% of Fractal’s revenue going forward. In FY25, it brought in $5 million in revenue for the company.
In FY24, Fractal spent approximately 7% of its revenue on R&D, focusing on generative AI, quantum algorithms and computational neuroscience.
The other agentic AI system being developed by the IPO-bound company is called ‘Project Pioneer,’ an agentic AI system allowing software development and data science tasks with minimal human intervention. “If you define the scope of a project, it does everything else. The system breaks it down into multiple roles, with each role performing their own job,” Velamakanni said, adding, “Results look very promising.”
Project Pioneer isn’t available yet, but Fractal expects to release it in “the next few weeks.”
Financial performance in FY24
In the last financial year, Fractal’s consolidated revenue stood at ₹2,196 crore, a 10.6% year-on-year increase from ₹1,985 crore in FY23. Earnings before interest, tax, depreciation and amortisation at the AI firm stood at ₹73.4 core in FY24, up from negative ₹188.4 crore in FY23.
However, the company reported a net loss of ₹54.7 crore in FY24, versus a net profit of ₹194 crore in FY23, according to the financials mentioned in the company’s annual report. The profit in FY23 was a result of the company giving up its 36.2% stake in Qure.ai at a value of ₹541 crore.
Fractal most recently raised $150 million in a pre-IPO funding round in January, Bloomberg reported, citing people familiar with the matter. The latest fundraise has bumped the company’s valuation to $2.4 billion, according to the report.
Fractal had previously raised $360 million in 2022. US-based private equity firm TPG led the investment, bumping it into unicorn territory, making it the first AI firm in India to reach such a valuation.
Last year, Bloomberg reported that Fractal was planning to raise $500 million through an IPO at a valuation of $3.5 billion. The company is expected to list on the bourses sometime this year.