Summary:
1. As AI scales, hybrid strategies and selective cloud repatriation are expected to accelerate in 2026.
2. Data readiness and inference will become a priority for organisations transitioning from AI experimentation to production.
3. KubeVirt is predicted to scale into mainstream production as an alternative approach to virtualised workloads.
Article:
In the rapidly evolving landscape of AI and data infrastructure, 2026 is poised to be a pivotal year for organisations seeking to build a solid foundation for long-term success. As regulators and boards scrutinize the whereabouts of sensitive data, Fred Lherault, Field CTO EMEA/Emerging Markets at Pure Storage, highlights the growing importance of hybrid strategies and selective cloud repatriation in the realm of AI.
The transition from AI experimentation to consistent, production-grade inference models is gaining momentum as enterprises move past the initial hype cycle and focus on operationalizing AI at scale. The shift in focus from training to inference underscores the critical need for a robust inference platform and data readiness for AI pipelines to thrive. As organisations grapple with the challenges of getting data AI-ready, streamlining and automating the entire data pipeline for AI is becoming a top priority.
Moreover, the intersection of AI and data sovereignty is reshaping cloud strategies and driving the acceleration of selective repatriation. Concerns surrounding data sovereignty and geopolitical tensions are prompting organisations to reconsider where their most crucial data is stored and how it can be accessed securely. This shift towards hybrid models positions AI-critical datasets and workloads closer to where they can be governed and controlled, without retreating entirely from the cloud.
In parallel, the rise of KubeVirt as a viable alternative for managing virtual machines through Kubernetes signifies a broader shift in infrastructure operations. The platform’s ability to unify virtualisation and containerisation needs aligns with the industry’s push towards a more streamlined and flexible operational model. As organisations seek to reduce complexity and avoid platform lock-in, the adoption of KubeVirt is expected to gain traction in 2026, alongside a growing demand for Kubernetes-native storage to support mixed-workload environments.
Overall, 2026 is projected to be a year of discipline, where organisations prioritize the operational foundation required for sustainable success. By transitioning from AI experimentation to production-grade models, modernizing data pipelines, reassessing cloud strategies with a focus on sovereignty, and exploring innovative virtualisation solutions like KubeVirt, organisations can position themselves for success in the ever-evolving landscape of AI and data infrastructure.
