Paytm Bets On AI To Drive FY27 Momentum

According to Paytm's own Q4 FY26 investor-relations blog post and reporting by Inc42, Founder and CEO Vijay Shekhar Sharma told investors on the company's May 2026 earnings call that Paytm will pursue inorganic growth opportunities but will limit all new investment to AI. Paytm reported FY26 revenue up 22% year-over-year to Rs 8,437 Cr and EBITDA improving by Rs 2,008 Cr year-over-year to Rs 502 Cr, its first full year of profit; Inc42 additionally reports a cash balance of Rs 13,315 Cr at the end of March 2026. Per the official blog, Sharma said: "We've built strong moats in payments and financial services by staying focused. Now, our next frontier is what we call marketing services which includes cloud, commerce, and AI-driven engagement." He described agentic AI as central to that push: "Agentic is a renewed opportunity for Paytm to gain market share across a number of categories. Seven times more people complete the funnel in an agentic workflow."

Per the official blog, Paytm is building applied AI models on top of open-source foundation models, including small language models tuned for the SMB context and optimized for voice and Indian languages. Inc42 reports that rather than build its own data centre, the company plans to rent data-centre capacity and run its models on top of that infrastructure, trading higher variable operating costs for faster deployment and access to specialized compute without operating on-premises cooling and networking at scale. The company is also embedding AI into merchant-facing hardware, including its Soundbox devices, which Paytm describes as evolving from a payment-confirmation tool into a small-business operating system.

Fintech firms reporting sustained profitability and large cash balances commonly accelerate AI-related experiments aimed at improving user-acquisition cost and monetization; Paytm's merchant GMV grew 27% year-over-year to Rs 6.5 lakh Cr and UPI GTV grew 46% year-over-year, per the official disclosure, giving the company a large existing transaction base to layer AI-driven personalization on top of. Outsourcing infrastructure while pursuing model-driven product improvements is a common strategy among capital-conscious platforms in the Indian fintech market.

Monitor whether Paytm or comparable fintechs publish technical details about model architectures, on-device inference approaches, or named partners for rented data-centre or GPU capacity; those disclosures would clarify latency, compliance, and cost tradeoffs behind an increasingly common rent-not-build infrastructure pattern.

Metrics tied to the AI initiatives, such as changes in customer-acquisition cost, merchant activation rates for AI-enabled devices, or incremental revenue from automated Paytm Money and Paytm Check-In flows across FY27; and whether Paytm names specific compute or model partners as its agentic AI push scales beyond the metrics disclosed so far.