India’s push for sovereign artificial intelligence is gathering momentum as domestic companies seek to build end-to-end capabilities across hardware, software and semiconductor ecosystems. Puneet Agarwal, CEO of VVDN Technologies, in an exclusive conversation with Bhaswati Guha Majumder, Associate Editor at CXO Media and APAC Media, outlines how the company is developing integrated, India-designed AI infrastructure solutions, advancing edge AI and vision technologies, partnering with global silicon firms, and positioning itself within the evolving semiconductor and data centre landscape to reduce external dependency and strengthen the country’s AI supply chain.
Tell us about the sovereign AI solutions VVDN is featuring at the IndiaAI Impact Summit and your expansion plans.
There are multiple perspectives. First, AI will be deployed across every enterprise, whether in factories, commercial spaces, or retail. Everywhere, AI use cases will be implemented. To deploy and implement these use cases, you need a strong backend infrastructure consisting of both hardware and software.
Until now, much of this hardware and software has been coming from outside India, where we have no control over the design or software. Therefore, there is a requirement for an Indian sovereign AI solution—a combined, integrated solution designed and manufactured in India.
The advantages include complete security and the ability to scale both hardware and software locally across industries. These solutions can also be deployed across local data centres. Over the next three years, there will be a huge demand and large-scale deployment of such solutions.
In terms of data centres, is VVDN in talks with players like Yotta and others?
A data centre can be understood in two or three ways. One is the physical infrastructure. For example, Yotta provides physical infrastructure involving compute, CPU, GPU, memory, storage, and more. Then there are applications that run on these data centres. For instance, fintech companies deploy banking and commercial applications, while oil and gas companies deploy their own specialised applications. So, a large ecosystem needs to be developed.
There are physical infrastructure providers like NTT Data and Yotta, and then there are solution providers, including software, hardware, and security companies, that use this infrastructure to deploy industry use cases.
We are in talks with many such companies, both within India and internationally. Our objective is to bring Make in India solutions to these companies. Currently, many are buying solutions from outside India. We aim to offer competitive alternatives so that India participates meaningfully in both physical infrastructure and solution layers.
Tell us about the Edge AI – Vision showcased at the IndiaAI Impact Summit Pavilion.
There are many solutions. On the vision and video side, we have developed a complete solution that includes edge devices like cameras and HD sensors that extract data. This data can be analysed directly at the edge, for example, within the camera itself.
Alternatively, the data can be sent to a local edge data centre, where multiple AI models process inputs from sensors and generate insights. We are building solutions that capture data, analyse it, deploy AI models, and generate inferences.
These applications include license plate recognition, face recognition, factory automation, and many more use cases. Once sensors capture the data, it can be analysed in multiple ways. This is how AI can be applied, whether for law enforcement agencies or for improving production throughput in factories.
VVDN has partnerships with several silicon companies. Could you elaborate on these and the supply chain ecosystem?
When designing systems, we acquire chipsets from companies such as Intel, NVIDIA, NXP, Qualcomm, and others. These companies provide silicon for different architectures. Based on industry use cases, we select the appropriate solution and partner to create a full system. That is the nature of our partnerships with silicon companies.
With ISM 2.0 now launched, the ecosystem of silicon development and manufacturing in India will be further localised. This is expected to be a major achievement. The initial programme was also successful, and a strong ecosystem is emerging.
We are now moving to advanced technologies such as 2-nanometre and 5-nanometre processes. India must build this supply chain to reduce dependency on global supply chains.
Over the past four months, memory prices have risen between two and ten times. India has little control over this because we lack a local supply chain. If we develop one, we can manage costs better. Otherwise, dependency will continue.
What are your expectations from the IndiaAI Impact Summit 2026?
This summit is a strong signal to the world that India will participate in AI from multiple angles. India aims to develop and build the entire AI ecosystem, including hardware and software solutions, and deploy AI at large scale domestically.
India is not only a development hub but also a massive target market. That is why leaders from around the world are visiting. They see India both as a technology development centre and as a major market.
Few countries have this dual advantage of scale and diverse demographics. We are in a strong position, and this summit sharpens the global focus on India’s leadership in AI.
