AWS recently held its AWS Summit 2025 in Bengaluru on May 7 and 8 at the KTPO Exhibition Center in Whitefield, bringing together over fifteen thousand attendees, including developers, partners and customer, to explore the advancements shaping India’s AI infrastructure.
Sandeep Dutta, President of India & South Asia, AWS, set the context by reflecting on India’s digital journey, pointing out the increase in internet penetration from 0.5% of the population 25 years ago to an estimated over 900 million users by the end of the current year.
He emphasised India’s capability to adapt complex technology for widespread access and national-scale transformation. Dutta cited examples like Prasar Bharti’s Wave OTT platform, which reached over 550 million users, and the Ministry of Women and Child Development’s Poshan Tracker app, which supports 100 million beneficiaries.
He spoke about the role of partnerships and customer trust in driving this transformation. Dutta mentioned the depth of trust earned and recognition for AWS’s culture, which includes principles of customer obsession and bias for action. This trust, he stated, is “powering some of India’s leading digital ambitions.”
He also highlighted the growth of Razorpay, India’s leading fintech firm, hosted on AWS servers. The startup handled $180 billion in transactions in May 2024 compared to $1 billion in 2017, reducing costs by 75%, he said.
India’s digital public infrastructure, such as Aadhaar, which has 1.3 billion digital identities, and UPI, which handles approximately 18 billion monthly transactions, was presented as a distinguishing factor.
Moreover, AWS reiterated its commitment to India, referencing its regional presence in Mumbai and Hyderabad, and a planned investment of $16.4 billion by 2030.
Dutta also announced that AWS will launch the AWS Marketplace in India later this year. The platform will include India-specific features, allowing users to interact with sellers in Indian rupees (INR). He added that it will support net banking, credit cards, and UPI.
In addition to technology, Dutta brought up sustainability, pointing out Amazon’s role as the world’s largest corporate buyer of renewable energy. This commitment is evident in India, with 1.3 gigawatts of renewable energy contracts and local initiatives like tree planting in Odisha.
AWS is reshaping businesses in the era of AI
Dutta characterised generative AI as “not another technology wave” but something that will fundamentally alter the paradigms of business operation.
Examples of early adoption in India include DigiYatra, which uses visual recognition to improve travel experiences, and Razorpay, which leverages Bedrock to create an AI assistant, Atlas, to accelerate content localisation, aiding its expansion to over 35 countries.
The discussion extended to the importance of culturally aware AI models.
On the second day of the summit, Dr. Pratyush Kumar, co-founder of Sarvam AI, one of AWS’s customers, shared his company’s focus on building AI exclusively for India, addressing the nuances of Indian languages and use cases. Kumar introduced the concept of “sovereign AI,” emphasising the need for strategic autonomy, data residency, and the ability for countries and eventually enterprises to build and optimise models internally.
Sarvam AI is working on foundational models for Indian languages and specific verticals like conversational AI and enterprise chat to drive widespread access to this technology for India’s large population.
From language models to developer tools, AI took centre stage across domains.
Abhijit Kane, co-founder of Postman, also shared how Postman leverages AWS services like container orchestration, Lambda, and Aurora to power their AI-driven API management platform, supporting their mission to enable connected developers.
Postman’s rapid growth has been closely tied to its evolving partnership with AWS. Kane explained that much of the company’s expansion happened as it moved from a single monolith to over 80 microservices, each managed by different teams. AWS played a crucial role in this shift by helping the company scale, control costs, and maintain high developer velocity.
He said that AWS helped Postman grow from 0.5 million to 40 million users and added that managing the nuanced interactions between microservices at that scale demands “a strong and mature cloud provider.”
Similarly, Murali Brahmadesam, CTO and head of engineering at Razorpay, discussed the company’s growth, AI-driven innovations, and how AWS has supported its scaling, including handling financial transactions and using AI to identify anomalies.
Satinder Pal Singh, Director, Solution Architects (India & South Asia), AWS, shared examples of Zepto, Angel One, and Amnic, highlighting how each has scaled operations, improved efficiency, or reduced costs using AWS services. He said Zepto grew from one dark store to hundreds across India in just three years and now handles millions of deliveries daily. Its entire tech stack is built natively on AWS, covering everything from supply chain optimisation to last-mile delivery.
Angel One, a leading player in the retail fintech space, has built a super app on AWS using a microservices architecture and container services. According to Singh, this has helped them cut application deployment times from days to minutes while reducing costs by over 18%.
The third example, Amnic, is a startup that has used AWS’s generative AI and identity capabilities to build a solution called FinOps.
The summit also featured a Generative AI Experience with hands-on demos, over 30 AI-focused breakout sessions, and live demos of tools like Amazon Q and PartyRock for building secure, responsible AI applications.
Building Blocks and Innovation at Scale
Sirish Chandrasekaran, vice president, Analytics, AWS, detailed the foundational building blocks that enable customers to innovate and scale. He made it clear that security is integrated in from the silicon all the way through the stack, keeping customer data safe. The scale is supported by AWS’s rapidly expanding global network.
He added that AWS operates the largest private network, with over 6 million kilometres of fibre optic cabling connecting its regions, availability zones and edge locations.
Chandrasekaran elaborated on compute options, including Amazon EC2 with over 850 instances catering to diverse workloads, from mission-critical enterprise applications to compute-intensive tasks like scientific modelling and machine learning.
Accelerated computing for AI workloads was highlighted with AWS’s partnership with NVIDIA and the upcoming P6 instances featuring NVIDIA’s Blackwell GPU.
Its investment in custom silicon was also presented as a key differentiator. Graviton processors, now in their fourth generation, offer performance improvements and significantly lower energy consumption compared to comparable x86-based CPUs.
Purpose-built AI chips like Inferentia and Tranium2 were discussed for their price performance benefits in AI training and inference, with Anthropic utilising Tranium2 for its Project Rainier compute cluster.
Singh announced that Tranium3 is coming soon.
Storage was another building block discussed, tracing the evolution of Amazon S3 as “storage for the internet.” With over 400 trillion objects, S3 supports various AWS services. Recent innovations include native support for Apache Iceberg table format, simplifying the management and querying of large analytical datasets, and S3 Metadata, enabling automatic creation and querying of system metadata and custom tags for objects.
Santanu Dutt, head of technology, Asia Pacific and Japan, commented on the evolution of AI, identifying the “Attention is all you need” paper, the Transformer architecture, and GPU capacity as key inflexion points that led to the recent acceleration in the field.
The summit conveyed a message of collective effort in driving India’s digital transformation. As Dutta quoted Ratan Tata, “If you want to walk fast, walk alone. But if you want to walk far, walk together.” This sentiment captured the spirit of partnership between AWS, its customers, and its developer ecosystem in shaping the future of India’s digital economy.
