What will it take for India to build AI for the world? How can startups balance AI-native innovation with AI-enabled growth, and what role should regulation and ethics play?
Questions like these and more found answers at the Google Cloud Startup Foundry Summit on September 19, 2025, at ITC, Gardenia, Bengaluru, held under the banner ‘Where India’s AI Ambition Meets Opportunity’. The event showcased India’s advancements in AI- through deep research, strong startups, ethical frameworks, and global ambitions.
Engineering the future with AI
Anand Rangarajan, Vice President of Engineering at Google DeepMind, began by painting a picture of AI not just as a tool, but as a force unlocking new frontiers.
Rangarajan spoke about AlphaFold, the system that has predicted over 200 million protein structures. “Before AlphaFold, a PhD student could spend years on one protein. Today, we can map millions,” he said, adding that this could lead to breakthroughs like enzymes that break down plastic.
He also introduced GraphCast, which gives highly detailed weather forecasts. “For farmers and planners, precise predictions are critical.”
On healthcare, he added that AI models in fields like dermatology and radiology are now “reaching or even surpassing human experts”.
Today, startups use tools like Gemini, Google’s multimodal AI, and Gemini Live for real-time mobile interactions. “The real challenge is not whether AI can do things,” said Rangarajan, adding, “It’s how we make sure it helps people at scale.”
AI in action: From potential to production
Later, Shai Alon, the Global Startup Evangelist from Google Cloud, showed how startups are using Google’s AI stack and what production looks like in real life.
“Successful startups don’t just build an AI product; they build their business around AI,” Alon said. He demonstrated tools and features, such as schedule-manipulating smart assistants like Pineapple that can adjust meetings to avoid traffic, showing how AI can go from concept to helping with daily tasks.
Alon also exhibited the AI Makeover app, which changes profile pictures and makes fun, themed images very quickly. He showcased how Gemini Text-to-Speech (TTS) works, and how the tool lets people talk to the app and get answers in real time.
Umesh Bude, CTO of Pocket Entertainment, described how Pocket FM is using AI to scale storytelling. The company, which has 200 million users to date, has built an in-house AI tool like Copilot, which helps writers maintain consistency across thousands of episodes, and Atlas, which localises stories into many languages.
“AI doesn’t replace humans, it accelerates creativity,” said Bude, whose comic studio, Blaze, now completes production much faster.
On the other hand, Akshat Mandloi, Co-founder of Samllest.ai, spoke about how the company uses AI in enterprise voice agents. He explained that voice is one of the oldest forms of communication, and the team had to build their own speech-to-text and text-to-speech systems to build voice agents that feel real and fast.
Mandloi emphasised that owning the full stack, both the models and deployment, is essential to get the speed and accuracy enterprises expect.
Building for the world from India
The evening’s highlight was a panel discussion titled ‘Building for the World from India’, led by Sandeep Kashyap, Head-Startups Business at Google Cloud, which explored the biggest question: What does it take to build a startup in India that succeeds globally?
Aakash Kumar, Managing Director at Z47, highlighted the importance of serving enterprise customers with clear problem statements. “Building a global B2B business is like solving a jigsaw puzzle,” Kumar said. “Every piece, product, trust, localisation, customer support, regulation, must fit perfectly.”
However, AfterShoot Founder and CEO, Harshit Dwivedi, shared that their early growth came mostly from users outside India. “At zero-to-one, survival depends on finding your first thousand true fans,” he added. He admitted that in India, handling things like local payments (for example, UPI) and localising properly had been challenging but critical.
Syed Shahrukh Ahmad, Co-founder, CloudSEK, also took the stage and described how different regions have different demands. Shahrukh explained that in the Middle East, customers want security and threat intelligence tailored to geopolitical risks, while in India, compliance and regulatory readiness often matter more. “Startups must understand these regional differences deeply, not assume what works in one place will work in another,” Ahmad says.
Responsible AI for India
The event continued with a fireside chat on ‘Responsible AI for India’ between Yolynd Lobo, Head of Public Policy and Government Relations at Google Cloud India, and Shweta Rajpal Kohli, President and CEO of Startup Policy Forum. The duo focused on the urgent need for ethical and policy frameworks to guide India’s AI growth responsibly.
“Regulation should be an enabler, not a barrier,” Kohli said, emphasising the need for clear guidelines for startups on responsible data use, transparency, and bias reduction to ensure the development of fair AI systems.
Lobo added that trust is fundamental for widespread AI adoption, adding that Google Cloud is working closely with government bodies and policy experts in India.
Google’s goal is to create guidelines that ensure AI technologies are safe, fair, and trusted by the public. Lobo stated that without public confidence, even the most powerful AI solutions risk failing to scale effectively.
Build with AI
In the closing address, the Foundry Summit hosted an impactful panel discussion on ‘Build with AI’, led by Naren Kachroo, Head of Artificial Intelligence, GTM India, Google Cloud India. The focus was on the Indian startups’ priorities, weighing AI-native products against AI-enabled solutions, while also examining where AI could have the biggest impact in India’s dynamic startup ecosystem.
Pratik Poddar, Partner at Nexus Venture Partners, highlighted the value in both AI-native and AI-enabled approaches. “Many are excited about AI-native startups built entirely around AI,” said Poddar, adding, “but established businesses using AI to improve products and operations can also succeed if they execute well.” He stated that the real priority is solving customer problems, not chasing headlines.
Prashant Parashar, SVP & Head of Technology at Delhivery, shared how AI helps their logistics operations scale efficiently, reduce waste, and improve speed. “AI tools don’t replace people; they enable us to serve customers better.”
Sahil Punjabi, Senior Director at Cloud Ambassadors, pointed out that the winners will be those who move AI beyond labs into real engineering workflows. Punjabi added that embedding AI into everyday tools and developer workflows makes it more powerful but less visible.
Vimal Kumar of Juspay offered a founder’s perspective, saying, “India has the talent and problems that demand AI-first thinking.” But his concern lies in whether startups can build responsibly, sustain their models, and maintain quality while scaling.
As the Startup Foundry Summit drew to a close, it was clear that the event had made a significant impact on India’s advancing AI journey, which is accelerating. From vision to execution, from policy to product, the pieces are falling into place. For founders, the message was loud and clear: build boldly and responsibly, think global, but stay rooted in local needs.
(Edited by Teja Lele)
Original source: in