Vinayak Desai

A very insightful compilation to spot the trend, moats, and where the next shifts may come from. The point on personalized UX through voice inputs really stood out, it is a space that will only get better with usage and feedback loops. Completely agree that trust must come before monetization. It opens the door, while value creation ensures scale. For AI startups, I believe incremental improvements, automation, or cost reduction alone won’t cut it. The real opportunity lies in rethinking how value is created, captured, and distributed with advancements in AI tools, infra, and LLMs. That’s where new economic value, and true disruption will emerge.

3

Had a great time being a panelist member at the Pre-Summit event of #IndiaAI Impact Summit 2026 on “Accessible by Design: A New Norm for AI” on 18 November at the India International Centre. A big thank-you to Chase Advisors and National Association for the Blind (Delhi) for convening such a timely and impactful discussion. My key take-aways: True accessibility must be embedded across the AI lifecycle - from dataset to interface to deployment - not just an add-on feature, not just a UI layer, and definitly not just a tick in a box. AI solutions must be built for all users, especially persons with disabilities, through inclusive design, voice enabled, multilingual-first capabilities and real-world constraints (low-bandwidth, low-end devices). Don't omit the people only then AI will serve every community. India has a unique opportunity to lead with #AI that is not only intelligent, but inherently inclusive by design. Collaboration matters: model builders, regulators, product teams, and disability advocacy communities need shared standards and joint accountability. Excited to take these conversations back into what we’re building at Sarvam AI — making accessibility not just a checklist, but a mindset baked into every stage. Looking forward to the full Summit in February, and to continuing the journey of building inclusive AI for all. #AI #Accessibility #Inclusion #IndiaAI #AIForAll #Sarvam #DEPwD

23

He co-built Mindtree from 0 to a listed company. Now he backs the next generation doing the same. Krishnakumar Natarajan, ex co-founder, Mindtree & managing partner, Mela Ventures — on what leadership actually looks like inside a scaling company.

171

10 comentarios

#BookRecommendation #ByteDance I am a staunch believer that the India technology ecosystem can benefit immensely from key learnings ("What to do?" and also, "what NOT to do?") from mature markets like China (one of the key reasons PROMAFT exists today!). 'Attention Factory', the book does a brilliant job of deeply researching the evolution of ByteDance - a company that has been rather obscure for it scale and impact. The book immaculately captures - 1. Thinking hard about the 'Consumer need' - Social network or Content Platform - using data to prove the same. 2. Evolution to 'Product-Market-Fit' - they were not the first to market! 3. Investing behind scalable technology - a recommendation system that 2x'ed' retention rate and usage in a matter of weeks 4. Rapidly scaling post 'Product-Market-Fit' - customer acquisition, M&A, offline activation - leaving gargantuan competitors like Tencent behind Super valuable for any entrepreneur looking to build & scale globally. Can't wait to re-read this again. Matthew Brennan

23

2 comentarios

India’s D2C space has seen a gold rush. Thousands of brands, millions in ad spend, and endless “shark” pitches. But here’s the not-so-sexy truth: very few are profitable. So, if you’re building a D2C brand in India, here are 5 things you must nail in the next 5 years. 1. Escape the Meta-Google Trap CAC is rising faster than my Swiggy bill on weekends. You can’t keep relying on paid ads to fuel growth. Community, content, retail presence, and referrals will separate survivors from casualties. 2. Make Retention Your Real CEO A 5% increase in repeat purchase rates can boost profits by up to 95% (according to a Bain study). Email, WhatsApp, loyalty, and cross-sell journeys are not “nice-to-have.” They’re the profit engine. 3. Omnichannel or Nothing Pure-play D2C is a myth in India. You need to be present where your customer shops - quick commerce, marketplaces, offline retail. Profitability often comes from the channel mix, not the D2C website alone. 4. Obsess Over Unit Economics, Not Vanity Metrics Stop chasing “valuation” screenshots. Focus on contribution margin, logistics cost, returns %, and repeat revenue. Because your investors might clap at top-line, but your P&L will scream at bottom-line. 5. Build a Brand Moat, Not Just a Product Products can be copied in 2 weeks in India. What can’t? Your brand story, trust, distribution, and community. If your only moat is “a better protein bar,” well… good luck. The next wave of Indian D2C will be less about blitzscaling and more about profit-scaling. The winners will be the ones who play a longer, more patient game. What’s one thing you think Indian D2C brands must fix before 2030? #D2C #EcommerceIndia #D2CBrands #Profitability #UnitEconomics #BrandBuilding #RetentionMarketing #Omnichannel #GrowthStrategy #StartupIndia #IndianStartups

45

5 comentarios

Swiggy just transformed on-demand commerce. By integrating the Model Context Protocol (MCP), Swiggy now allows users to order food and groceries directly through AI agents like ChatGPT and Google Gemini. No more app scrolling—just natural conversation. Why this shifts the paradigm: Zero Friction: Complex tasks become simple prompts. Active AI: MCP turns chatbots into transaction engines. New Interface: AI is becoming the primary storefront. This evolution highlights the massive potential of the Google Cloud ecosystem. The AnoCloud Role: We architect the cloud foundations—secure APIs, data pipelines, and scalable infrastructure—that make these advanced AI integrations possible.

3

I was pleasantly surprised to see a TAP agent perform so well. In the example shared, you can see how the agent creates a neatly formatted report analyzing the impact of various groups based on the Budget released by India. The TAP agent created a decent report including the Indian flag colors on the title page and personalized it based on what was known about the user. When tried with best model and mode in ChatGPT, it output a dismal looking PDF with formatting completely butchered. It was barely readable.

6