Cloudera has launched Anywhere Cloud, a data and AI platform created to run AI applications across hybrid cloud environments without requiring enterprises to relocate their data.
The platform gives companies a single control plane for deploying and managing data and AI services across these environments. It can operate as a standalone platform or work with existing enterprise data lakes.
Cloudera’s approach addresses a major obstacle facing companies moving AI projects into production: enterprise data is typically spread across multiple locations, while security rules, regulations and costs can limit where that data can be moved. Research cited by Cloudera found that 73% of IT leaders say infrastructure performance limitations have interfered with operational initiatives.
Anywhere Cloud is designed to allow enterprises to bring AI compute to the data instead of transferring large data sets to the infrastructure running the application. This enables a company to place a workload in a public cloud, sovereign environment or its own data center based on regulatory, business or infrastructure requirements.
For Cloudera, the strategy seems focused on making mixed, hybrid infrastructure an advantage rather than an obstacle. Instead of requiring companies to consolidate data in one cloud environment before deploying AI, Anywhere Cloud allows compute and AI services to operate where the data already resides. Agentic AI
This model could be particularly relevant for companies deploying agentic AI, which can require access to large volumes of proprietary corporate information. Keeping data in its existing location can reduce the security and governance issues created when sensitive information is copied between environments.
Cloudera is also building automation into the platform through an agentic copilot. Users can issue instructions in plain language to perform tasks like deploying compute engines and creating data workflows across connected environments.
The company is combining this automation with centralized zero-trust governance. According to Cloudera, applications deployed through Anywhere Cloud maintain enterprise policies governing areas like data access and compliance.
The platform uses Apache Iceberg, the Polaris catalog and common APIs to connect analytics engines and data. Cloudera has also added GPU acceleration for Apache Spark 4.1 in Cloudera Data Engineering. The capability uses NVIDIA’s cuDF CUDA-X library and is built to speed up Spark workloads without needing changes to PySpark or SQL code. For enterprises prepping large data sets for AI applications, GPU acceleration could shorten processing times for data engineering.