Indian IT ‘Agentic AI’ Mode: ON

Indian IT 'Agentic AI' Mode: ON THE BELAMY WEEKLY NEWSLETTER OF AIM Monday, Jan 20, 2025 | By Amit Raja Naik A few days ago, in a conference call with journalists and analysts, the CEO of a top Indian IT company was asked a simple question—How much revenue is coming from AI? His answer was a long explanation about AI projects, workforce training, and client interest. But no number. This has become a pattern. While TCS, Infosys, Wipro, HCLTech, LTIMindtree, and Tech Mahindra are all pushing AI harder than ever, they continue to remain tightlipped about how much it's actually making them. Contrast this with Accenture ($4.2 billion in generative AI bookings and $1.4 billion in generative AI sales since September 2023) or Genpact openly reporting AI-driven growth, claiming that data-tech-AI revenue grew 9% YoY to $569 million, making up 47% of total revenue. It also reported a 60% increase in GenAI solutions in production (now 130+).  The divide is unmistakable. But, the good news is that a majority of Indian IT firms are moving beyond developing generative AI chatbots and implementing RAG solutions to capitalise on the advancements in agentic AI. Before we dive into the AI buzz from Indian IT this quarter, let's look at some of the Top Stories of the Week: The hype around AGI and OpenAI's o3 is dying down as scrutiny mounts over its FrontierMath benchmark claims. It turns out OpenAI had access to much of the test data—and even funded it—without full disclosure. "We made a mistake in not being more transparent," admitted Tamay Besiroglu, associate director and co-founder at AI research institute, EpochAI. AI experts like Gary Marcus are comparing this to the Theranos scandal.  Silicon Valley's tech wars are getting uglier. Mark Zuckerberg recently slammed Apple for profiting off app developers, while Salesforce's Marc Benioff called Microsoft Copilot a "branding disaster". Meanwhile, Sam Altman labelled Elon Musk a "bully," accusing him of wanting control over OpenAI, while Musk fired back over OpenAI's shift to a for-profit model. Why so much contempt? Read the full story here. India's GCC boom is getting bigger while the GCCs get smaller. Now, even one-person GCCs are a possibility, claims ANSR chief Lalit Ahuja. With over 825 GCCs already operating and another 620 expected by 2030, the landscape is shifting from massive corporate hubs to nano and micro GCCs—some as small as 10-50 employees. Read the full story here. So, What are Indian IT Firms (Really) Saying?  Coming back to Indian IT, each company has its unique way of pitching AI. TCS: AI is Now Business as Usual TCS has embedded AI into a majority of its transformation deals and sees rapid growth in AI adoption. "AI adoption is no longer an experiment; it is becoming core to business strategies," said its chief, K Krithivasan. "We are now past the initial chatbot phase. AI is moving into automation and intelligence."  Translation: TCS is integrating AI deeply but isn't breaking out AI revenue separately. Infosys: AI is in Every Deal Infosys is using small language models and multi-agent AI to automate processes and improve efficiency. "Generative AI is now a part of everything we do—any large program, transformation, or cost-efficiency initiative," said Infosys chief Salil Parekh. "We are moving from AI pilots to full-scale enterprise adoption." Translation: Infosys is betting on AI, but like TCS, it won't break out AI revenue. Wipro: AI is a Workforce Play Wipro is focusing on reskilling employees and building AI-driven consulting. "AI will be a net positive for us and the industry," said its CEO Srini Pallia. "We categorise AI projects as AI-led, AI-infused, and AI-powered solutions." Translation: Wipro sees AI as a long-term play, but it's still in its early days. HCLTech: AI Through Cloud Partnerships HCLTech is partnering with AWS, Google, and NVIDIA to expand its AI offerings. "We continue to see strong demand for AI and GenAI solutions, especially through our AI Force and AI Labs," said its CEO Vijayakumar C. "Our cloud partnerships are helping us scale AI faster."  Translation: HCLTech is leaning on cloud players to expand AI services. LTIMindtree: AI for Everyone LTIMindtree is making AI accessible across industries. "Our AI strategy is delivering results, with major AI wins in manufacturing, finance, and energy," said CEO Debashis Chatterjee. "We have trained our workforce to ensure AI fluency across the company."  Translation: LTIMindtree is betting on broad AI adoption rather than niche solutions. Tech Mahindra: AI with a Twist—Building Its Own Models Tech Mahindra is one of the few Indian IT firms that are building their own AI models rather than relying on OpenAI or Google. "We are the only SI building sovereign AI models from scratch—this gives us a unique market position," said chief Mohit Joshi. "Agentic AI will redefine enterprise AI, moving beyond chatbots to autonomous workflows." Translation: Tech Mahindra wants to be different, but it's unclear how much revenue these efforts will bring. AIM VIDEO: Is it about passion or just endless hustle?  While leaders debate productivity vs burnout, one question remains—How long can you really stare at your screen before it starts staring back? Catch the latest episode of DECODED, where our tech journalists break down L&T chairman SN Subrahmanyan's "90-hour" remark. Where's the AI Revenue?  Meanwhile, despite all the AI talk, Indian IT firms resist breaking down their AI revenue separately.  Here's what they're saying (or not saying): "We do not disclose AI-specific revenue numbers, but AI engagements are growing," said N Ganapathy Subramaniam, chief operating officer of TCS. He added that AI is part of larger IT contracts, making it hard to separate. Similarly, Infosys chief Salil Parekh said, "Counting AI revenue is not a measure we would like to go after because it misrepresents the impact of AI." He added that AI is an enabler, not a standalone product. "Every single dollar of revenue will have a generative AI component embedded in it," said Debashis Chatterjee, CEO and managing director of LTIMindtree, implying that AI is everywhere, so breaking out numbers is unnecessary.  Unlike Accenture and Genpact, which openly report AI revenue, Indian IT is cautious—perhaps because AI revenue is still small or they don't want to set expectations too high. On the bright side, Indian IT seems more confident than before. This is evident in the kind of deals they are closing and the work they've been doing to bring generative AI and, now, agentic AI to life.  "We are seeing many enterprises move beyond the hype of AI-washing and focus on real, measurable business impact from AI adoption. Our approach is to bring practical AI solutions that integrate seamlessly into core business processes, rather than just showcasing AI for the sake of it," said Parekh, subtly critiquing competitors who aggressively market AI without clear impact. Just a quick reminder before we wrap up—MLDS 2025 is almost here! Get ready for three days of AI, innovation, and hands-on learning. Don't miss out. Reserve your spot now! Winners & Losers: Who's Leading the AI Race? Surprisingly, there are no losers in FY25 Q3—every major IT firm is investing in AI. Tech Mahindra, TCS, and Infosys lead with proprietary models, deep AI integration, and multi-agent solutions, while HCLTech, Wipro, and LTIMindtree are making solid AI strides.  The real challenge? AI revenue transparency. In other news,  Google DeepMind introduced "Inference-Time Scaling for Diffusion Models," showing that allocating more compute during inference can dramatically improve output quality. Qualcomm and MapmyIndia teamed up to bring AI-powered, affordable vehicle connectivity to Indian automakers.  Samsung has partnered with Eka Care to launch the 'Health Records' feature in the Samsung Health app.  The US has finally opened its Bengaluru consulate, strengthening Indo-US tech and business ties while promising visa services soon. Sakana.ai unveiled Transformer2, a self-adaptive AI model that dynamically adjusts its weights for different tasks, marking a shift from static to lifelong learning AI systems. IIT Madras professor Balaraman Ravindran has been elected as an AAAI Fellow, recognising his pioneering contributions to AI research and education in India, particularly in deep reinforcement learning. Microsoft introduced MatterGen, a generative AI tool for materials discovery, capable of designing entirely new compounds with unprecedented precision. Karya launched Karya Institute, an AI think-tank focused on creating ethical digital work opportunities for marginalised communities.  For Brand collaborations, write to info@aimmediahouse.com You received this email because you signed up to the updates from AIM. Click here to unsubscribe if you do not want to receive emails from us.