In a country as linguistically diverse and digitally fragmented as India, the promise of artificial intelligence has often felt aspirational rather than accessible. While global AI platforms compete on scale, parameters, and benchmarks, one Delhi-based startup is quietly asking a more grounded question: What does AI look like when it’s built for how India actually communicates?
That question defines Puch.ai—a voice-first, multilingual AI platform designed to reach Indians through the simplest interfaces they already use: WhatsApp and phone calls.
The Founder’s Vision: AI for Everyday India
Puch.ai was founded in 2025 by Siddharth Bhatia, an AI researcher and entrepreneur with a strong academic and industry background. A BITS Pilani graduate, Siddharth earned his PhD from the National University of Singapore under the President’s Graduate Fellowship and worked with teams at Google Research and AWS. Before Puch, he built TurboML, focusing on continual learning systems.
Despite the option to remain in global tech roles, Siddharth returned to India with a clear belief:
AI adoption in India would not be driven by dashboards or prompts, but by language, voice, and distribution.
He was joined by Arjit Jain, an IIT Bombay alumnus, who brings deep engineering expertise to the platform. Together, they set out to build what they describe as AI for people who’ve never used AI before.
The Product: AI Where India Already Is
Puch.ai doesn’t ask users to download an app or learn how to “prompt.” Instead, users simply message or call 9090909090 on WhatsApp and start interacting instantly. The platform supports 22 Indic languages, handling natural accents, colloquial speech, and voice notes—areas where many global models still struggle.
This approach has led to a defining metric: nearly 60% of Puch’s users are first-time AI adopters. From students and homemakers to small business owners, users rely on Puch for education support, health queries, local fact-checking, creative content, and everyday problem-solving.
Under the hood, Puch runs on a self-hosted, sovereign AI stack, avoiding reliance on foreign APIs or external data pipelines. This ensures stronger privacy control, higher reliability, and alignment with India’s evolving data and AI governance landscape.
Adoption and Impact: Distribution Over Brute Force
Within months of launch, Puch.ai scaled to close to one million monthly users, largely through organic adoption. Its growth reinforces a familiar Indian tech lesson: distribution beats sophistication.
With WhatsApp penetration exceeding 75% in India, Puch meets users where they are—especially in Tier 2 and Tier 3 towns, where language barriers and app fatigue often limit adoption. Voice interaction further lowers the barrier, making AI feel conversational rather than technical.
Funding and a Different Growth Playbook
Unlike most AI startups in a capital-heavy ecosystem, Puch.ai is bootstrapped. The company has chosen to prioritize product-market fit, distribution, and user trust over rapid fundraising. While it has drawn interest and public appreciation from industry observers and policymakers, Puch continues to operate lean, focusing on scale through usage rather than spend.
Its hiring strategy reflects the same philosophy. Puch recruits through open, high-paying remote internships with no formal degree requirements, even welcoming high-school students. Performance, not pedigree, determines growth—challenging conventional startup hiring norms.
India’s AI Moment, Reimagined
As global cloud providers and venture funds pour billions into India’s AI ecosystem, Puch.ai represents a parallel path—one shaped by language, access, and cultural context.
It’s too early to predict how big Puch will become. But the pattern is familiar. India doesn’t just consume technology. It adapts it—and often reshapes it entirely.
Sometimes, the future of AI doesn’t arrive through an app update.
It arrives through a simple message. Or a missed call.
