From Traffic to Food Wastage: How AI is Solving India's Modern Problems "When credit was widely used, no one foresaw the 2008 fallout. Sustainability is a similarly critical issue," said Sameer Phukan, Director and Head Customer and Partner Engineering, Intel India

India, a land of immense diversity, is grappling with a range of modern challenges in its metro cities, including pollution, inadequate infrastructure, traffic congestion, and higher crime rates. To address these emerging issues, Indian leaders are increasingly turning to Artificial Intelligence (AI).

Sharing insights on how AI can aid in solving the problems of 1.4 billion Indians, Sameer Phukan, Director and Head of Customer and Partner Engineering, Intel, emphasized the need for inclusivity. "India is a country with multiple dialects and languages. Ensuring we have datasets and intelligent data available for each of these local languages is key," Phukan noted.

He also stressed the importance of indigenous AI models, adding, "The problems we face today, like traffic management or ordering services, are very different from those in the U.S. Indic models that rely on local data will make AI solutions more pervasive across the broader population."

AI for India, By India

For a non-English-speaking population, leveraging AI poses unique challenges, both for users and for AI companies trying to cater to their needs. Addressing this, Subramanian Natarajan, Director of Customer Engineering at Google Cloud, explained their efforts in developing models that reflect India's linguistic, cultural, and socio-economic diversity.

"To build and train models on datasets that represent not just language diversity but also gender, ethnicity, and socio-economic strata became a major challenge for us," Natarajan shared. However, he highlighted, "Today, whether we talk about Hindi, English, Hinglish, or even mixing multiple languages with emotions, the out-of-the-box accuracy levels are astounding. Domain-specific training further improves these levels dramatically." For Natarajan, India's diversity represents a unique opportunity: "Solve for India, and you solve for the world."

AI's Impact on Indian Businesses

Inclusivity also extends to accessibility, a point emphasized by Anoop Balakrishnan, Head of AI Product Management at BigBasket. Before the AI boom, the company explored building self-checkout stores tailored for Indian consumers. Traditional self-checkout systems, which require extensive hardware to track customers and products, were expensive and complex. Instead, BigBasket opted for a simpler, cost-effective model powered by computer vision.

"We trained our system on around 10,000 FMCG SKUs and 250 fruits and vegetables. Customers can now walk into a store and check out independently," Balakrishnan explained. Inspired by ATMs, the model uses AI to track purchased items, delivering a seamless shopping experience.

As BigBasket is planning to expand with this model into Tier-III and Tier-IV cities, Balakrishnan acknowledged initial trust issues among consumers regarding such systems but cited data to dispel concerns: "People are not here to steal everything. While there are some cases, self-checkout systems work in India and have great potential."

Driving Sustainability with AI

While addressing how companies can minimize the carbon footprints and power consumption Natarajan says Google Cloud provides tools to help companies using Generative AI measure the energy consumption of their workloads directly in the console. These tools allow users to make sustainable choices by selecting alternative hardware platforms that may consume less energy, with a little bit of compromise on either latency or performance. "Depending on your sustainability threshold, you should be in a position to choose the right kind of platform. This is the starting point," he stressed.

Adding to this, Phukan compared the current sustainability challenges to the U.S. credit card crisis of 2008: "When credit was widely used, no one foresaw the 2008 fallout. Sustainability is a similarly critical issue."

Balakrishnan shared how AI helped BigBasket address food wastage and greenhouse gas emissions. Using solar power and electric vehicle fleets for deliveries, the company also implemented AI-driven predictive systems to forecast demand. "Our AI predicts demand in specific areas, like near a drugstore, and auto-indent orders to warehouses. This has drastically reduced food wastage," he explained. However, challenges remain, particularly with over-ordering of FMCG and perishable goods.

All speakers were speaking at a Global Conclave of AI.