Evolving the cloud-native foundation for the next generation of AI
Microsoft outlines why cloud-native platforms must evolve to support AI agents, emphasizing coordinated compute, storage and networking, open technologies and measurable business outcomes.
AI agents introduce new demands on cloud infrastructure, requiring systems that adapt dynamically to real-time decisions, long-running processes and cross-environment coordination. Unlike traditional applications, agents gather intelligence, use tools and change course based on evolving conditions, pushing platforms to deliver reliability, visibility and responsiveness. The shift calls for infrastructure that evolves alongside AI capabilities rather than requiring a complete rebuild of existing cloud-native foundations.
Microsoft argues that compute, storage and networking must operate as a unified system to meet the fluctuating demands of AI workloads. The company highlights the need for application platforms that simplify building and running AI workloads while preserving flexibility through open technologies. Decisions made today about cloud-native foundations will determine how effectively organizations can deploy AI in the future, with outcomes measured by the useful work enabled.
Basecamp Research used Azure to assemble 9.7 trillion nucleotides for its therapeutic-design models, validating results across three applications within three weeks by coordinating compute, data and platform capabilities. The company emphasized the importance of integrating scaled data, model scaling and compute scaling, crediting Microsoft for its expertise in these areas. This example illustrates how AI infrastructure must be designed around specific outcomes rather than simply scaled up.
Microsoft notes that 95% of new AI agent deployments will use Kubernetes by 2028, according to Gartner, underscoring the role of established cloud-native components in AI adoption. Azure Kubernetes Service is presented as a way to access open-source innovation while reducing the complexity and risk of managing it in production. The company positions its platform as a bridge between open technologies and practical business deployment, aiming to make advanced capabilities accessible without requiring specialized expertise.