Fujitsu and Rackspace Technology announced new AI partnerships on September 10, 2026, focusing on the systems businesses need to move artificial intelligence from trials into daily operations. Fujitsu signed a new agreement with Palantir Technologies Japan KK covering Palantir AIP and Palantir Foundry, while Rackspace joined the NVIDIA Cloud Partner Program.
The announcements point to the same business challenge from different angles: AI requires more than a model or a powerful processor. Companies also need reliable data, secure infrastructure, engineering support, and teams that can keep the full system running after deployment.
Fujitsu expands its Palantir partnership
Fujitsu agreed to serve as a Global Forward Deployed Engineering, or FDE, Partner for Palantir Technologies Inc. Its partnership with Palantir has been in place since 2020, and the new agreement expands Fujitsu’s role around Foundry and AIP.
Fujitsu is making a significant investment in building Forward Deployed Engineering capabilities with Palantir. The company will also bring its own AI technologies, including the Takane large language model, to support customers adopting Foundry and AIP.
The goal is not limited to installing software. Fujitsu will support the development of custom AI applications within sovereign, production-grade architectures, giving customers a path to build systems around their own operational needs and control requirements.
“This partnership is about bringing Palantir’s sovereign AI architecture to Fujitsu’s customers across Japan,” Kevin Kawasaki, Global Head of Business Development at Palantir Technologies, said.
Yoshinami Takahashi, Corporate Executive Officer, Corporate Vice President and COO of Fujitsu, is also part of the leadership behind the agreement. The arrangement builds on work Fujitsu has already completed for a leading Japanese manufacturer.
A supply chain project shows the business case
Fujitsu implemented a supply chain resilience solution built on the Palantir platform for that manufacturer. The project integrated data across more than 3,000 suppliers and 18 factories, connecting information from a network large enough to make fragmented decision-making difficult.
Within one year, the manufacturer achieved more than $10 million in cost savings. It also doubled its operational productivity and significantly accelerated its response to disruptions and its decision-making.
That example gives the Fujitsu agreement a clear focus. Foundry and AIP are being positioned as tools for custom business applications, while Fujitsu supplies the engineering capabilities needed to fit those tools into a sovereign, production-grade environment.
Rackspace combines NVIDIA systems with Palantir platforms
Rackspace Technology said it joined the NVIDIA Cloud Partner Program on September 10, 2026. The relationship advances Rackspace’s AI Cloud strategy and brings NVIDIA technologies across all layers of the enterprise AI stack.
Those capabilities include Compute, Networking, Visualization, and NVIDIA Enterprise Software. Rackspace will combine them with its engineering and operations to help customers move AI into production across private and hybrid environments.
Rackspace’s Institutional Sovereign Pod architecture is designed to give enterprises control over their data, models, and AI environments. The Sovereign Pod will bring together NVIDIA Blackwell architecture with Palantir Foundry and AIP, creating an environment operated by Rackspace through managed operations and forward-deployed engineering.
The setup is designed to support AI deployment at scale while maintaining sovereignty, security, and operational control. Rackspace will support deployment choices based on data residency, governance, security, and compliance requirements, which can shape where AI systems run and how organizations manage them.
Rackspace’s role extends beyond providing infrastructure. The company says it will make the entire AI system work and remain accountable for the outcome, bringing together compute, cloud infrastructure, data, resilience, and AI platforms.
“Enterprise AI is becoming production infrastructure. At that point, the question is no longer simply who provides the GPU. It is who is accountable for operating the complete system,” Gajen Kandiah, CEO of Rackspace Technology, said.
The push from AI experiments to working systems
Rackspace’s agreement with NVIDIA and its Sovereign Pod architecture reflect a broader shift in how organizations approach enterprise AI. The focus is moving toward complete systems that can run under real operating conditions, rather than isolated demonstrations built around a single model or processor.
Raj Mirpuri, NVIDIA’s vice president of global AI clouds and infrastructure ecosystem, said, “AI is moving from experimentation to production, and enterprises and nations need full-stack AI factories that can operate continuously, securely and at scale.”
Mirpuri also said NVIDIA and Rackspace can help more organizations put AI into production securely, efficiently, and at scale. Rackspace will combine NVIDIA accelerated computing with its own engineering and operations to help customers maintain control of data, models, and intelligence.
Together, the announcements place engineering and accountability at the center of enterprise AI. Fujitsu is expanding its ability to develop custom applications with Palantir, while Rackspace is building an operated environment that connects NVIDIA infrastructure with Palantir platforms. Both approaches focus on turning AI capabilities into systems businesses can control and use in production.
Based on
