Infosys posted strong financial results for Q3 FY25, highlighting growth in revenue, operating margin, and free cash flow. CEO and MD Salil Parekh attributed the performance to the company’s strong positioning in digital services and its growing enterprise AI capabilities, particularly in generative AI and AI agents.
“In generative AI, we have built four small language models for banking, IT operations, cyber security and broadly for enterprises in generative AI,” Parekh said. He added that the company is developing over 100 new agents for its clients, many of which are already in use.
“We are very clear about what we are doing with generative AI,” Parekh said. He added that Infosys is not doing ‘AI washing’ as others in the industry might be and, instead, doing real generative AI work.
“We have several discussions with clients where they would like to use the small language models that we have built…They are built by using the proprietary data that we have. Some clients are asking us to build a small language model of their own.”
For AI agents, Parekh explained that Infosys has built a research agent for its clients which is not just a PoC. “They are now using that in their product area to support how queries are looked at, and where their own people and their own customers can use this agent,” Parekh said, adding that these agents are able to reduce the work of 18 days to eight days.
The company reported a revenue of $4.94 billion, reflecting a sequential growth of 6.1% year-on-year (YoY) and 1.7% quarter-on-quarter (QoQ) in constant currency terms.
The global IT giant reported a net profit of $804 million for the quarter, which demonstrates robust financial performance.
“We continue to strengthen our enterprise AI capabilities, particularly focusing on generative AI, which is witnessing increasing client traction,” Parekh further said. “This has led to another quarter of strong large deal wins and an improved deal pipeline, giving us greater confidence as we look ahead.”
The company secured $2.5 billion in large deal total contract value (TCV) with 63% net new deals, growing 57% sequentially.
Headcount also increased for the second consecutive quarter, currently standing at 3,23,379. “We have had a strong hiring in Q3 with the addition of over 5,000 employees,” Parekh added. Infosys is on track to onboard 15,000 to 20,000 freshers at the group level in FY25.
Indian IT Reporting Strong Generative AI Pipeline
When Accenture reported its earnings for the first quarter of fiscal 2025 (ending November 30, 2024), the company set a record of $1.2 billion in generative AI bookings, reflecting growing client investments in this space.
In FY24, the company reported $4.2 billion in GenAI bookings overall. It was expected that Indian IT firms would follow suit. However, even TCS, L&T, HCLTech, and now Infosys, in all of their recent earnings calls, shied away from revealing the revenue from generative AI, citing several reasons like “AI is now part of every deal”. While all of them said that they are working on GenAI and AI agents, the exact details remain unknown.
For example, in its latest Q3 FY25 earnings call, TCS reported that its clients are actively investing in generative AI and agentic AI while building robust data foundations. K Krithivasan, CEO and MD, said that TCS has been working on a drug discovery project with a client who was able to identify 1,300 molecules and further filter them to 12 molecules.
Moreover, HCLTech announced during its Q3 FY25 earnings call that it is advancing its GenAI strategy with an aim to integrate AI services into 100 clients by FY26.
Similarly, in the last quarter, Infosys finally revealed that it is working on small language models, building multi-agent frameworks, and doing incredible work with generative AI. The exact numbers gained from this work, however, remained undisclosed.
Parekh added that the company believes small language models will provide clients with a powerful tool, allow them to build business logic on top of it and unlock new potential. “This combination will form the foundation of the small language model, which is being tailored for different industry applications,” he added.
Infosys is Doing a Lot of Generative AI
Just a few hours before the result, Infosys had launched a suite of AI-driven features for the Australian Open (AO) 2025 in partnership with Tennis Australia. As per the company’s press release, this marks a major milestone in their seven-year collaboration, which aims to enhance the tennis experience through technology.
Last month, Infosys also launched a Google Cloud Centre of Excellence at its Bengaluru campus to drive enterprise AI innovations. The initiative, powered by Infosys Topaz, aims to help businesses use generative AI for transformative growth.
Furthermore, as promised earlier, Infosys is finally becoming an AI-first company. For instance, at the Microsoft Building AI Companions for India event in Bengaluru, Infosys CTO Rafee Tarafdar said that developers have been using GitHub Copilot for over a year, which now has about 20,000 users generating nearly a million lines of code every few weeks.
Beyond internal work, at Meta’s Build with AI Summit held in Bengaluru last year, Infosys announced a partnership with Meta to utilise the Llama stack, a collection of open-source large language models and tools, to build AI solutions across industries.
As an early adopter of Llama 3.1 and 3.2 models, Infosys is integrating these models with the in-house AI platform Infosys Topaz to create tools that deliver business value. One example of such a tool is a document assistant powered by Llama that improves the efficiency of contract reviews.
