New Delhi, June 10 -- Cyber warfare has emerged as a new and critical frontier in modern geopolitical disruptions and tensions. The same was the case with the recent India-Pakistan strike on the back of the Pahalgam attack.

The military conflict between the two countries that lasted less than a week saw the unleashing of massive cyber attacks. A report in early May showed that Maharashtra Cyber, the infosec division under the government of Maharashtra, records over 10 lakh online attacks targeting India.

Against this backdrop, there is a growing imperative to build private, secure digital infrastructures. As Artificial Intelligence (AI) increasingly powers mission-critical defense systems, ranging from intrusion detection and remote weapon stations to real-time language translation, the security and integrity of these technologies have become paramount for national defense agencies.

To this end, Bengaluru-based AI services provider Kogo AI's co-founder and chief product officer (CPO) Praveer Kochhar, advocates for private AI, especially for defense establishments, that doesn't need public or hybrid cloud to operate.

Private AI

Private AI infrastructure is an approach where customers can leverage AI without having to compromise the control of their data, while ensuring privacy and compliance. It brings AI models closer to customer data.

According to Kochhar, with AI models shrinking in size and becoming faster and more efficient, the world might see a shift in the next 12-24 months with enterprise hosting their own AI infrastructure, complete with GPUs, models, agents, and workflows, within their private cloud or on-premise setups.

"This approach is particularly crucial for industries dealing with sensitive data, such as banking, insurance, healthcare, defense, and the public sector. Private AI allows them to own not just their data, but also the models, training processes, agents, and the resulting IP. Without that ownership, building defensible, domain-specific AI capabilities becomes nearly impossible," said Kochhar in an interview with TechCircle.

Kogo AI, which offers a low-code platform for building AI agents, partnered with chipmaker Qualcomm to provide a private AI stack for enterprises and institutions. Through this partnership, Kogo AI is also co-developing an Agentic AI solution for the Indian Army.

"We are currently working on integrating with select systems operated by the Indian Army to build AI-powered agents that support faster and more effective decision-making. These systems collect vast amounts of intelligence from a variety of sources, including sensors, surveillance platforms, and operational databases," said Kocchar, adding that their goal is to add an agentic AI layer that helps analysts rapidly interpret and act on this data.

He added that in this six-month-long engagement, Kogo AI will be developing 'research-driven' solutions from the ground up. Kochhar said that while this project is 'rewarding', few challenges persist.

"It's an extremely challenging environment, primarily because we have to operate entirely within their infrastructure. No data can leave their premises, and our platform must be deployed inside their secure environment.

Every customisation or modification has to be done on-site, which naturally extends development timelines. That said, we've found ways to navigate these constraints as they arise."

What is next for Kogo AI

Kogo AI has partnered with system integrators such as Tech Mahindra and Coforge. The Tech Mahindra partnership was announced in May, aimed at jointly developing and deploying enterprise-grade AI solutions. While Kogo AI will develop infrastructure for agent-based AI systems, Tech Mahindra will offer consulting and digital transformation services.

Kogo AI was founded in 2018 and was initially focused on building AI solutions for the travel and mobility sector. About a year and a half ago, the company made a strategic decision to offer horizontal AI solutions, thereby serving diverse sectors. Kochhar called it a natural evolution of their capabilities.

We developed a platform where any industry can build and deploy AI agents, automation workflows, and custom solutions. There's strong demand for agentic automation, but the supply side is still fragmented and overly complex. Our goal is to simplify that, and that's exactly where we're seeing strong traction," he added.

Mapping company MapMyIndia currently owns a 40% stake in Kogo AI, acquired in two separate tranches in 2022 and 2024. In the next two quarters, Kogo AI will announce fresh investment, Kochhar said.

"We're performing well on the revenue front, and while we're awaiting a few deal closures, we expect to share some concrete revenue milestones soon. There's strong momentum, and we're confident that we'll emerge as one of the standout AI success stories from India." While India is Kogo AI's main focus market, it is also expanding its reach to the US and the Middle East.

Published by HT Digital Content Services with permission from TechCircle.