Amid the AI buzz, startups must focus on India’s ‘real big fundamental issues’: Google’s Preeti Lobana

Google India Country Manager Preeti Lobana believes startups should tap into the power of artificial intelligence to solve real-world problems in healthcare, education, and sustainability sectors—essentially those areas that mirror India’s deeply rooted challenges.

“My advice is that… if you are truly talking about succeeding in a big way over a long period of time, I think it’s critical to focus on the real big fundamental issues that exist in India,” Lobana said during a fireside chat with Subrata Mitra, Partner, Accel, at Google I/O Connect India 2025 in Bengaluru on Wednesday.

While healthcare in major metros is largely adequate, she emphasised that the shortage of professionals and infrastructure in Tier II and III cities, as well as rural areas, remains a challenge, and making innovative solutions in these regions will be very helpful.

Lobana shared the example of Cloudphysician, which has built an AI-powered platform called RADAR to enable smart ICUs with 24/7 remote monitoring. By partnering with Google Cloud, the healthcare startup has implemented its solution in 200 hospitals across 22 cities, treated over 1 lakh critically ill patients, and reduced mortality rate by 40%.

Enterprises are using Google Cloud to build secure, scalable, AI-powered infrastructure based on foundational models like Gemini, enabling them to reimagine their core processes, which, according to Lobana, is like “compressing a decade of transformation in a couple of years”.

Given India’s large, multilingual, and socio-economically diverse youth population, education is another critical area where AI can drive meaningful impact.

“If we can build solutions that are adaptive, customised for each individual for their particular context, that will really help us bridge the skilling gap that exists,” Lobana remarked.

Google is conducting advanced research in Indic languages—a capability demonstrated earlier in the keynote. Manish Jain, Research Engineer at Google DeepMind, also showcased Project Astra, a research prototype of a universal AI assistant. 

Jain highlighted that millions of Indian students study in English, which is not always their most comfortable language, creating a barrier to learning. His demo showed how students can interact in languages like Hindi to get learning support from an AI tutor that understands both their curriculum and their language.

“Imagine the impact for every individual, anywhere, having access to this personal multilingual assistant in education and multiple other domains,” Jain said, after sharing an early look at the work underway in Google’s labs.

Speaking in this context, Lobana said the goal is to collaborate with “many of you [startups] to come up with AI solutions that will ensure AI is truly universally accessible for Indians.”

On the topic of climate change and sustainability, she noted that water is becoming an increasingly scarce resource, and expressed confidence that AI can help detect leaks, improve resource management, and optimise renewable energy grids.

She also pointed out the importance of leveraging AI, alongside India’s Digital Public Infrastructure (DPI) in these sectors. “A lot of these DPI solutions are inherently frugal and scalable, and will work beautifully with many other countries. So, grab the dual opportunity of building in India for India, but also for the world,” Lobana added.

Privacy and security by design

According to Lobana, digital India thrives on trust. “All of this that we are talking about just goes to nothing if the trust is eroded,” she said, suggesting that safety and security should be built into the architecture from day one, not added later, so users aren’t burdened with making complex security decisions.

As cybersecurity threats are becoming more sophisticated, she added that it’s not just about security, but also about privacy by design, ensuring strong data protection measures, and that any data sharing preserves user privacy.

Stressing the importance of having clear design principles when building AI, the India head noted that Google follows its own AI principles on safety, security, privacy, and accountability, ensuring continuous testing and rigorous red teaming to identify vulnerabilities before they can be exploited.

According to her, when automating critical processes, it’s essential to include human intervention to prompt before sensitive actions are taken, as “transparency and provenance” are vital for ensuring data integrity and building trust.

Google embeds an invisible digital watermark in all AI-generated content, including text, images, audio, and video, which remains intact even if the data is shared or modified.

Accel’s Mitra noted that in a world where LLMs already have access to vast public data, startups must go beyond. To truly stand out, they need access to proprietary data, usually from enterprises.

However, gaining that access is challenging, so startups must build trust—either directly or through partners—he explained, citing startup Emma, which has partnered with consulting firms to earn enterprise credibility.

Mitra underscored the importance of flexibility, noting that when a startup co-designs with an enterprise and asks for their data, the enterprise will expect something in return. “Building large companies or building from a startup to a large company is usually about being flexible,” he said.

At the third edition of Google I/O Connect India, the company presented its latest AI innovations tailored for India’s developer ecosystem, including on-device intelligence, agentic capabilities, and localised AI, aimed at enabling developers to contribute to India's role in the global AI landscape.

India has the second-largest number of active developers on Google Play globally, contributing over 1 million developer jobs in 2024.