India’s Startup Titans on Building, Surviving & Winning | Stream 100
This is the 100th live stream episode of T1, featuring founders and executives from five Indian startups: Servify (device lifecycle management), Square Yards (proptech), UpGrad (edtech), Shadowfax (logistics), and InMobi (adtech/consumer internet). Each guest discusses their company's business model, growth journey, competitive strategy, and views on AI and the future of their respective industries.
Summary
The episode opens with Sri (Sridhar), founder and CEO of Servify, explaining how his company acts as an invisible platform engine behind major device brands like Apple and Samsung, managing the entire lifecycle of over 700 million devices across 40+ countries. Servify handles extended warranties, trade-ins, repairs, diagnostics, and affordability programs without running physical service centers, instead integrating with OEM networks and acting as a neutral technology intermediary. Sri describes the complexity of enterprise sales cycles (12-18 months to go live), the challenge of integrating with dozens of disparate systems across global brands, and how remote diagnostics and fraud detection technology (SF Shield equivalent) help distinguish genuine claims from gaming the system. He also shares the origin story of Servify's Apple partnership, tracing it back to a cold email he sent to Apple's head of care in 2009.
Kanika of Square Yards discusses how the company built India's largest proptech platform by addressing the opacity and fragmentation of the Indian real estate market. Square Yards covers search, transaction, home financing, and interior renovation, with 2080 crores in revenue and 175 crores EBITDA, operating at the intersection of growth, scale, and profitability. Kanika explains how the company built trust on both the buyer and developer sides through transparency tools, verified listings, and 3D walkthroughs, and how AI is being used to make property search conversational. The company is preparing for an IPO and sees India's long-term urbanization story as a one-way door for residential real estate demand.
Ronnie Screwvala of UpGrad and RSVP reflects on his entrepreneurial journey from UTV/media to education and deep tech investing. He describes UpGrad's consolidation strategy as a 'learner journey consolidation' rather than corporate consolidation, covering the full education lifecycle from school to mid-career upskilling. Ronnie discusses the importance of being counterintuitive, the challenges of underemployment versus unemployment in India, and his views on AI's role in augmenting (not replacing) the workforce. He also shares perspectives on his Swades Foundation, his RSVP film production ventures, and his 50 million deep tech fund with investments in companies like Digantra (space tech).
Vaibhav of Shadowfax (listed logistics company) explains how the company built foundational logistics infrastructure — from address geocoding and route optimization to fraud detection (SF Shield) and proprietary mapping (SF Maps) — that now enables quick commerce at scale. He describes how misroute rates dropped from 7-8% to 1.25% using their own maps trained on Indian address patterns, and how their return quality-check model uses LLMs to verify product returns at the doorstep. Vaibhav also discusses the evolution from 5-day e-commerce to 10-minute quick commerce, the role of EV logistics in reducing costs, and the future of agentic commerce in logistics.
Finally, Mohit and Abhay of InMobi — India's first unicorn — discuss the company's 20-year journey from mobile advertising to building Glance, an agentic commerce platform now live in the US. InMobi processes $2.5 billion in media across 3 billion consumers and serves 2 trillion ads per month across 170+ countries. The co-founders emphasize the importance of staying focused (mobile-first even when US advertisers pushed for omnichannel), building systems that scale without human dependency, and competing directly with Google and Meta by going deep in a niche rather than broad. Glance AI is positioned as a personal shopping concierge trained on 300 million users, solving for shopping inspiration rather than just efficiency.
The episode closes with the hosts reflecting on 280 days and 100 episodes of T1, discussing the preparation behind each stream, the inspiration from TBPN, and their mission to create an ESPN-for-tech equivalent for India's startup ecosystem.
Key Insights
- Sri argues that Servify's commercial model is tied to customer NPS, meaning the platform is financially penalized for bad service experiences — unlike retailers who can blame the product and walk away.
- Sri claims that enterprise sales cycles for global OEM partnerships (like Apple or Samsung) take 12-18 months just to go live, and that one large computer brand contracted in 2020 had still not gone live by June 2026 due to multi-country integration complexity.
- Sri asserts that extended warranty claim rates are in single-digit percentages annually (6-8%), but that the pricing model is based on failure frequency and average cost of service during a claim, making it actuarially sound.
- Sri states that retailers make disproportionately more profit from extended warranties than from device sales — earning ~4,000 rupees on a 9,000 rupee extended warranty versus ~2,000 rupees on a 1 lakh rupee iPhone — and that it operates as a negative working capital business for retailers.
- Kanika argues that Square Yards' entry point was not technology but behavioral change — building transparency tools that made buyers informed rather than dependent on brokers, and providing developers with a reliable, professional sales channel.
- Kanika claims India needs to build 50 million homes over the next two decades but currently builds under 1 million per year, framing this gap not as a crisis but as the foundational opportunity for Square Yards.
- Ronnie Screwvala argues that UpGrad's M&A activity is not corporate consolidation but 'learner journey consolidation' — assembling the full X and Y axis of education from school-age to mid-career across price points and specializations.
- Ronnie contends that underemployment is sometimes more frustrating than unemployment, because a person who has invested heavily in education but is paid one-fourth of their potential is in a worse psychological position than someone without a job.
- Ronnie claims that Indian IT companies' arbitrage model is not a benchmark for AI leadership, citing Sunil Mittal's observation that IT companies spend more on share buybacks than investing in the future.
- Vaibhav states that Shadowfax's proprietary SF Maps reduced misroute rates from 7-8% to 1.25% by training on Indian address patterns, including phonetic transliteration issues and landmark-based addresses like 'Ramu Paan wale ke paas wali dukaan'.
- Vaibhav argues that quick commerce is impossible to run manually — the entire allocation, routing, and demand-surge management must be algorithmically automated, drawing on his prior experience designing high-frequency trading algorithms.
- Vaibhav claims that Shadowfax's return-verification LLM model can identify correct product, brand, and size even from a folded garment image, and that errors in reverse logistics pickups directly impact profitability at scale.
- Abhay of InMobi argues that subscription-based models funding AI experiences have a 'natural ceiling' and cannot fund the capex investments being made in AI infrastructure, making advertising and agentic commerce the two scalable monetization models for AI surfaces.
- Mohit of InMobi argues that LLMs represent only 5% of the engineering problem in AI systems — the remaining 95% is classical software engineering: token optimization, input sanitization, output moderation, and rendering at scale.
- Abhay of InMobi contends that developing for India is orders of magnitude harder than developing for the US due to linguistic, regional, and infrastructure complexity, meaning founders who can build for India can more easily build for global markets than the reverse.
Topics
Full transcript available for MurmurCast members
Sign Up to Access