Puch AI has spent much of its short life in the glare of public attention—an unusual trajectory for a startup that isn’t even a year old. In a matter of months, the Bengaluru-based company has moved from audacious, headline-grabbing gestures—such as a largely symbolic $50-billion bid for Perplexity AI and Google Chrome—to signing a Rs 25,000-crore memorandum of understanding (MoU) with the Uttar Pradesh government. At the centre of it all are co-founder and CEO Siddharth Bhatia, a technologist with a research bent, and co-founder and CTO Arjit Jain, as they attempt to carve out a place in India’s fast-evolving AI landscape.
Distribution-First
Founded in June 2025, Puch AI is built on a clear, if contrarian, thesis: that the next wave of AI adoption in India will come not from power users, but from those still on the margins of the digital economy. Its product reflects that belief. Instead of nudging users towards yet another app, it embeds AI into familiar interfaces such as WhatsApp and voice calls—lowering the barrier for those uncomfortable with typing or navigating new platforms. Its systems are tuned for Indian languages and accents, with use cases spanning everyday assistance, education, healthcare and content creation.
“The business model seems to be to create a utility use case that can be plugged into platforms and services as AI adoption rises. We’ve seen this with AI-enabled note-takers, for instance. Typically, the goal is to either move to a licensed model or exit when valuations become attractive,” said Pareekh Jain, founder and chief executive of tech advisory firm EIIRTrend.
This distribution-first strategy—prioritising access over building proprietary models—has defined both Puch AI’s positioning and the scrutiny it attracts. That scrutiny sharpened after the announcement of its MoU with the Uttar Pradesh government, which laid out an ambitious blueprint: AI parks, large-scale data centre infrastructure, a citizen-facing “AI Commons” platform, and even an AI university. The scale—pegged at Rs 25,000 crore—ensured instant visibility.
It also raised eyebrows. Online commentary quickly pointed to the gap between the ambition of the proposal and the company’s early-stage profile, including its limited operating history and the absence of publicly disclosed financials.
Data from market intelligence platform Tracxn places Puch AI in a crowded but still-forming segment of conversational AI assistants, particularly those built around messaging platforms like WhatsApp. It ranks fifth among 27 active global competitors, alongside players such as Luzia and Zapia, which have raised significant funding and operate similar models.
While several of its peers have secured between $287,000 and $50 million in funding, Puch AI is listed as unfunded, with no disclosed equity backing. Yet, it is estimated to hold a striking 43% market share within a tracked subset—ahead of competitors such as faff (30%) and Luzia (13%)—underscoring the early, fluid nature of this category.
The Uttar Pradesh government later clarified that the agreement is preliminary and non-binding, with any forward movement contingent on due diligence, project evaluation and regulatory approvals.
Bhatia, in a detailed public response, sought to address the concerns. He said the MoU is structured as a public-private partnership that does not involve taxpayer funding, with investments to be brought in through external partners and rolled out in phases. He also pushed back against widely circulated claims about the company’s revenue, calling them a data error.
“For now, as a private company, we’re not required to make our valuation or revenue public. But if you’re curious, we’re not bootstrapped—we’re a well-funded startup,” he said.
Puch AI’s approach—built on open-source systems and distributed through widely used communication platforms—comes with its own set of challenges. Chief among them is platform dependency, particularly its ongoing friction with Meta over WhatsApp policies.
Scaling Through Friction
At the heart of the issue are WhatsApp’s evolving rules for third-party AI chatbots. Puch AI has suggested that access restrictions and platform controls tilt the playing field in favour of Meta’s in-house AI, constraining how independent services can scale. The standoff highlights a broader fault line in the AI ecosystem: the growing tension between startups building on closed platforms and the gatekeepers that control them.