GlobalLogic sees higher conversion rate of AI PoCs to deployment

GlobalLogic says 40-50% of its AI pilots move to production, far above global averages, as enterprises focus on agentic and physical AI for real business gains

Even as doubts are raised about the ability of artificial intelligence to create value for organisations, digital engineering-focused firm GlobalLogic is seeing a materially higher conversion of AI proof-of-concepts (PoCs) into full-scale deployments than what global surveys suggest. The company has seen 40–50 per cent of its AI pilots moving into production, as enterprises increasingly focus on agentic and physical AI to drive measurable business outcomes. 

“In our case, the success rate is much higher than what some of the broader industry reports indicate,” said Ethan Matyas, chief delivery officer at GlobalLogic, in an interaction with Business Standard. 

“Roughly 40–50 per cent of the PoCs we work on turn into follow-on production work.” 

The comments come amid growing scepticism around enterprise AI returns, following reports such as the MIT-led State of AI in Business study, which found that while nearly all companies are experimenting with AI, only a small fraction are seeing tangible value at scale.

 

Matyas attributed GlobalLogic’s higher conversion rate to upfront cost modelling, change management planning, and clearer expectations around AI economics. “A lot of disappointment stems from not factoring in things like LLM costs or organisational change. We spend time coaching clients even before the PoC so there are no negative surprises later,” he said. 

Piyush Jha, GVP and Head APAC at GlobalLogic, said AI is already reshaping delivery economics across engineering, quality, and operations.

 

“Almost all the customers with whom we have done PoCs, we have moved stuff to production. So we’re actively using AI. We’re using AI in production setups at like 112 different places,” Jha said.

 

Jha also added that the early movers are clearly seeing benefits of efficiency-led gains. He also added that AI at scale is also gaining ground. “If earlier a programme required 15,000 people and today the same outcomes can be achieved with 12,000 — while redeploying the remaining talent toward growth initiatives — that is AI at scale. Similarly, when organisations begin to decouple revenue growth from headcount expansion, it reflects a fundamental shift driven by AI adoption,” he said.

 

Matyas added: “When we look at product and platform development, the question is not only whether we can build the same asset with 20–30 per cent less time or fewer people — because the answer to that is yes. The more strategic shift is how organisations are redeploying those efficiencies. What we consistently observe is that clients have sizeable innovation backlogs that they have been unable to fund. The real opportunity unlocked by AI is the ability to channel newly created efficiencies into building new products, accelerating roadmaps, and retiring long-standing technical debt.”

 

While generative AI continues to dominate current deployments, both executives said the next phase of growth will be driven by agentic and physical AI.

 

“With agentic AI, executives, developers, architects, and even HR professionals will have agents that actively perform tasks, not just generate content. That’s where efficiencies of 40–70 per cent become real,” said Jha.

 

GlobalLogic has also set up a physical AI lab at its Noida facility over the past six months, focusing on use cases that combine AI, IoT, and edge computing. One such deployment involves reimagining building management systems to optimise energy usage, space management, access control, and intelligent parking.

 

“Physical AI is becoming mainstream far faster than we expected, largely because compute is now available at the edge,” Jha said. Deployments are already underway in the US and Japan, with India expected to follow over the next 12 months, he added.

 

Matyas said advances in edge computing, coupled with cloud and hyperscaler ecosystems, are enabling entirely new classes of applications. GlobalLogic is a launch partner for Google’s agentic AI framework and has ongoing collaborations with Nvidia, including joint engagements at the parent Hitachi level.

 

India now accounts for nearly half of GlobalLogic’s global workforce and plays a disproportionate role in its AI innovation, according to the company.

 

“India is not just a cost arbitrage for us — it is an innovation arbitrage,” Jha said. “Nearly two-thirds of our AI demos, accelerators, and content are built by teams in India.”

 

The company has also seen strong growth in its global capability centre (GCC) business, with revenues from GCC clients growing at around 60 per cent year-on-year for the past three years. “In the last three years, our GCC business has grown nearly 2.5–3x, and we don’t see that slowing down,” Jha said.

 

Contrary to concerns that GCC expansion could hurt IT services firms, Jha said GlobalLogic sees them as partners. “GCCs today are innovation drivers, shipping end-to-end products. Our experience with high-end digital transformation makes us a natural partner for them.”