Kyndryl expands Google Cloud service for AI workloads

Kyndryl has expanded its distributed cloud services with Google Cloud, focusing on application modernisation and AI-ready workloads across on-premises, edge and public cloud environments.

The expanded offering combines Google Distributed Cloud with Kubernetes-based application modernisation on Google Kubernetes Engine. Kyndryl will provide consulting, implementation and managed services to help customers build and run distributed cloud environments across hybrid and multicloud infrastructure.

The announcement comes as large organisations face growing pressure over where data is stored and processed, particularly as they adopt more AI tools and cloud services. Businesses in regulated industries increasingly want to keep some workloads closer to their own sites or within specific jurisdictions while still using cloud software and management tools.

The service is aimed at enterprises managing fragmented technology estates, rising costs and tighter regulatory requirements. Demand is also growing for systems that can support data-intensive and AI-driven workloads across multiple locations while meeting data sovereignty rules.

Distributed model

Under the arrangement, Kyndryl will help customers deploy Google Distributed Cloud alongside Kubernetes-based application modernisation. The goal is to let organisations run cloud-native applications in locations that fit their regulatory, latency and operational needs, while retaining control over where data sits and how it is governed.

The service is also designed to provide a more consistent operating model across private cloud, on-premises data centres and public cloud environments. It standardises governance, security and lifecycle management, while allowing workloads to be placed or moved as requirements change.

One focus area is the use of containerisation and Kubernetes to update older applications. Kyndryl said tools such as Gemini Enterprise can assist in that process and reduce the complexity of modernisation work.

The company highlighted two main use cases: speeding up application modernisation through Kubernetes and containerisation, and supporting data-heavy and AI-related workloads closer to where data is generated. That reflects a broader shift among enterprises that want to process information at the edge or within their own facilities rather than rely solely on centralised public cloud regions.

Partner comments

Kyndryl described the expansion as a response to customers seeking greater oversight of complex cloud estates.

"As data and AI workloads scale, customers are looking for greater control and visibility across their cloud environments," said Giovanni Carraro, Global Strategic Alliances Leader at Kyndryl. "Together with Google Cloud, we're helping enterprises modernise applications and operate more effectively across distributed environments - without compromising performance or compliance."

Google Cloud said the partnership is aimed at customers whose needs cannot be met by a public-cloud-only model.

"Google Distributed Cloud extends Google Cloud infrastructure, advanced AI and services directly into customer environments," said Eliot Danner, Managing Director of Google Distributed Cloud. "Together with Kyndryl, we're enabling organisations to run applications where public cloud alone cannot meet customers' regulatory, latency, or operational requirements."

Market pressure

The announcement reflects a broader shift in how enterprises approach cloud architecture. Earlier phases of cloud adoption often involved centralising workloads in hyperscale public cloud platforms, but that model has proved harder to apply where data residency laws, operational resilience requirements or network delays shape system design.

For companies handling sensitive records, industrial data or low-latency services, distributed cloud models have become more relevant. These approaches allow cloud software and management frameworks to run inside customer-controlled environments, at edge locations or in dedicated infrastructure, rather than only in remote provider-operated regions.

Kyndryl, which focuses on infrastructure and managed services, has positioned itself as a partner for large organisations operating these mixed environments. Its work with Google Cloud adds to broader efforts by service providers and cloud vendors to tie AI deployment more closely to infrastructure location, governance and application redesign.

That is becoming more important as AI projects move from experimentation into production. Running these workloads often raises questions about where training or inference data is held, whether data can cross borders, and how companies maintain operational consistency across multiple computing environments.

The expanded service is intended to address those issues by extending cloud operations into customer environments through a common management approach. The aim is to let enterprises run applications in locations where public cloud alone cannot meet regulatory, latency or operational requirements.