Cloudera has formalized a two-track strategy that combines six years of support for Cloudera 7.3.2 with parallel deployment of its new AI and data platform, Cloudera Anywhere Cloud. Announced on the 20th (local time) at EVOLVE26 Singapore in Singapore, the approach allows large production workloads to remain on existing systems while new AI and data applications run on Anywhere Cloud. Customers using Cloudera 7.1.9 do not have to upgrade to 7.3.2 first, because both versions were designed to be compatible with the new platform. Anywhere Cloud uses a modular architecture that lets enterprises deploy or upgrade services independently, including running Spark 3 workloads while Spark 4 is introduced separately. Cloudera also plans to support its own and external processing engines and models across public clouds, enterprise data centers and sovereign infrastructure. Its AI Factory strategy focuses on enterprise systems and inference rather than model training alone, with sovereign inference (running models where data is stored) intended to reduce data-transfer costs and improve control. The company cited a real financial institution's KYC workflow analysis and said its inference service can improve speed by up to 40% in certain environments for the same model. Cloudera is also expanding Field Deployment Engineer support, industry-specific AI blueprints and hybrid deployment capabilities to help customers move from experimentation to production while preserving existing investments.