Paytm Partners with Groq to Accelerate AI-Driven Financial Services Innovation

Paytm's AI Workloads Now Powered by Groq | AIM

Paytm will use GroqCloud to support its 'ongoing work in building high-performance AI models' for enhanced transaction processing, risk assessment, fraud detection, and consumer engagement across its platform. India's payments giant Paytm has announced a partnership with Groq, a US-based company that develops specialised hardware for AI inference. Paytm will use GroqCloud to support its 'ongoing work in building high-performance AI models' for enhanced transaction processing, risk assessment, fraud detection, and consumer engagement across its platform. GroqCloud is Groq's cloud-based service for developers and enterprises to run AI inference -- the process of deriving outputs and insights from a trained model. The service is powered by Groq's proprietary Language Processing Units (LPUs), which are custom processors explicitly built for inference, delivering significantly faster performance and higher energy efficiency than traditional GPU-based systems. Narendra Singh Yadav, chief business officer at Paytm, said, "We have been steadily advancing our AI capabilities to make payments faster, more reliable, and deeply intelligent." "This collaboration with Groq strengthens our technology foundation by enabling real-time AI inference at scale. It marks another step in our journey to build India's most advanced AI-driven payment and financial services platforms." Currently, Paytm is leveraging AI for both consumer-facing and internal operations. Recently, the company launched what it calls India's first AI-powered Soundbox, a business device aimed at small and medium enterprises. The AI Soundbox comes with an in-built assistant that can talk to merchants in 11 Indian languages, offering real-time insights and responses based on their payment activity and business data. Merchants can ask the device for summaries of their collections, trends, and performance, turning it into what Paytm describes as an "AI business assistant." In a live demo at the Global Fintech Fest this year, Paytm CEO Vijay Shekhar Sharma asked the soundbox about the amount of money it received on the soundbox in a single day ,and it replied within a few seconds in the same language. A report from the Economic Times this year stated that the company is using AI-driven quality checks for onboarding merchants, image and text recognition (OCR), name matching and deduction of merchant category code (MCC) from storefront photos. Besides, its in-house platforms, such as Paytm ARMS (a merchant lifecycle insights platform) and Paytm Pi (a fraud and risk detection system), leverage AI to automate merchant onboarding, fraud detection, segmentation, and pricing optimisation. The report further added that the company uses conversational AI agents to provide personalised assistance to customers via text and voice across 11 languages. Earlier this year, Perplexity, the AI-powered search engine tool, was also made available on the Paytm app on both Android and iPhone. Users can ask their queries, to which the tool will provide responses with citations on a Perplexity interface.

Paytm to deploy GroqCloud to enhance AI on its platforms

Paytm is partnering with Groq, a US-based AI leader. This collaboration will bring ultra-fast AI technology to Paytm's platform. The aim is to significantly improve transaction processing, risk assessment, and fraud detection. Customer engagement will also be enhanced. This move signifies Paytm's commitment to building India's most advanced AI-driven financial services. Fintech firm One 97 Communications, which owns Paytm brand, will deploy Groq AI technology to build high-performance AI models that enhance transaction processing, risk assessment, fraud detection, and customer engagement across its platform, the company said on Wednesday. Founded by a former Google engineer, Jonathan Ross, US-based Groq specialises in ultra-fast AI chips called language processing units. "Paytm (One 97 Communications Limited)...announced a partnership with Groq, the U.S.-based leader in real-time AI inference, to bring fast, intelligent, and cost-efficient AI to its platform. Under this collaboration, Paytm and its associate entities will deploy GroqCloud, powered by Groq's purpose-built LPU, to achieve significantly faster, more cost-efficient and scalable AI inference compared to conventional GPU-based systems," the statement said. The fintech firm said that the advancement will support its ongoing work in building high-performance AI models that enhance transaction processing, risk assessment, fraud detection, and customer engagement across its platform. "This collaboration with Groq strengthens our technology foundation by enabling real-time AI inference at scale. It marks another step in our journey to build India's most advanced AI-driven payment and financial services platforms," Paytm chief business officer Narendra Singh Yadav said. The partnership aims at enhancing Paytm's ability to deliver instant, intelligent, and reliable experiences across its payments and financial services ecosystem. "Groq is proud to support Paytm in driving real-time AI innovation at national scale. Core to our mission is delivering broad compute capacity to serve the world's biggest problems which AI will uniquely solve. Paytm's ambition closely aligns with our own to make AI useful and accessible," Groq, GM of APAC, Scott Albin said.

Paytm Q2FY26 Earnings Call - CEO Says AI Is Revenue Multiplier

Artificial intelligence (AI) will unlock new revenue streams for Paytm as the fintech company increasingly integrates the technology in its core business operations, said Founder and CEO Vijay Shekhar Sharma during the Q2FY26 earnings call. "AI is a revenue line item. It brings newer service, newer business...phenomenally more number of things that we can do," Sharma remarked. He clarified that AI is not merely a cost-cutting tool for the company, but is also contributing in creating new products and services. "We want to ensure that AI is a part of our business model as much as possible. Till now, we've been putting it into the cost side, efficiency side. Now, we are trying to build AI on the product side and feature side, which was not possible earlier," said the CEO. Sharma further stated that the company is building AI agents for merchants, which will serve as digital "chief operating officer, chief finance officer, or chief marketing officer" for smaller businesses. "Typically, large enterprises get third-party companies like Snowflake or Accenture. We plan to do it for small businesses in the country," he said. The Paytm founder also pointed out that the company is already piloting AI-based subscription services for merchants, with plans to later expand it into an "AI commerce cloud" line item in its financial reporting. "We do believe that one year forward, there should be a commerce cloud line item. The cloud element could be the AI element that will start to go back once again," he said. Further, Sharma expects that Paytm Soundbox will evolve into a smart and interactive device in the future. "I've been saying this, Soundbox will sound like a feature phone some days later. I mean, this will be like a smartphone," Paytm's Founder said during the Q2FY26 call. Interestingly, the CEO also revealed that Paytm is internally testing some AI products that it plans to roll out to larger businesses, mid-sized retailers, or even e-commerce companies like Flipkart and Swiggy in the coming quarters. The CEO's remarks came after Paytm reported a 24% year-on-year (YoY) in its operating revenue to Rs 2,061 crore in Q2FY26, helped by an increase in merchant subscriptions, higher payments gross merchandise value (GMV), and robust growth in financial services distribution. Despite an uptick in its top line, the company reported a sharp 98% YoY drop in its consolidated net profit to Rs 21 crore during the quarter, mainly due to a one-time exceptional loss of Rs 190 crore incurred on account of its gaming joint venture First Games Technology. In its earnings release, Paytm said that its indirect expenses (including ESOP costs) were down materially in Q2FY26, primarily on account of AI-led efficiency. Meanwhile, indirect costs, comprising marketing expenses and cloud and software expenses, among others, decreased 18% YoY to Rs 1,064 crore during the quarter under review. In a separate exchange filing today, Paytm said that it has teamed up with US-based AI company Groq to power real-time AI for payments and platform intelligence. The fintech says that this partnership is expected to enhance its ability to deliver "instant, intelligent, and reliable" experiences across Paytm's payments and financial services ecosystem. As per the exchange filing, Paytm and its associate entities will deploy GroqCloud to achieve significantly faster, more cost-efficient and scalable AI inference compared to conventional GPU-based systems. Notably, Paytm says that this move is part of ongoing efforts to build high-performance AI models that boost transaction processing, risk assessment, fraud detection, and customer engagement. Commenting on this move, Narendra Singh Yadav, Chief Business Officer (CBO) at Paytm, said, "We have been steadily advancing our AI capabilities to make payments faster, more reliable, and deeply intelligent. "This collaboration with Groq strengthens our technology foundation by enabling real-time AI inference at scale. [And] it marks another step in our journey to build India's most advanced AI-driven payment and financial services platforms." In line with Yadav's remarks, Paytm has recently been betting big on AI and is increasingly adopting the technology across business verticals. For instance, last month, Paytm's wealth tech arm Paytm Money partnered with JioBlackRock to launch a new AI-powered active equity fund.