Cloudera has announced Cloudera Anywhere Cloud, a new platform for taking data applications and AI into production on hybrid infrastructure. It can deploy services across public clouds, sovereign infrastructures, and private data centers from a common control plane, with a focus on a growing enterprise-AI problem: using data where it’s stored without having to move it to a specific cloud.
Cloudera Anywhere Cloud in 30 seconds
- Cloudera Anywhere Cloud runs data and AI services across public clouds, sovereign infrastructures, and private data centers.
- It aims to avoid data migrations for certain AI use cases while keeping governance centralized.
- It integrates open technologies like Apache Iceberg and supports engines such as Spark, Kafka, and Trino.
- An agent-based copilot can turn natural-language instructions into data operations and infrastructure tasks.
- Cloudera especially targets private, sovereign AI and agent-based applications with it.
The launch comes as many organizations try to move past early AI experiments. Building a proof of concept with an API and a small document set is fairly simple. Putting that same system into production, where data is spread across multiple clouds, internal systems, and legacy platforms, is a lot harder.
Cloudera argues this fragmentation is holding enterprise projects back. It cites its own research that 73% of surveyed IT leaders think infrastructure performance limits have hurt their operational initiatives. That figure comes from Cloudera’s studies and should be read in the context of the launch.
Enterprise AI is running into the data reality
Anywhere Cloud rests on a simple idea: in many large organizations, the data AI needs isn’t in one place and can’t be moved freely.
A company might keep some data in a public cloud, other data in its own data center, and certain datasets under specific residency, privacy, or regulatory requirements.
Cloudera’s architecture aims to lay a common layer over these environments. Organizations can deploy data and AI services independently and choose where to run each workload based on technical, economic, or regulatory criteria.
That’s especially relevant for so-called sovereign AI, where the model choice, execution location, data residency, and infrastructure control all matter.
Cloudera says certain applications can run directly on existing repositories without moving or copying data. It presents this as an alternative to the large data migrations that modernization projects traditionally involve.
It’s not that any application runs anywhere unchanged. Whether it works depends on the engine, the architecture, the compute needs, and the data sources. Anywhere Cloud is the layer Cloudera wants to use to simplify distributing and managing these workloads.
Apache Iceberg, Spark, Kafka, and Trino keep the open piece
The technology choices stand out too.
Cloudera Anywhere Cloud uses Apache Iceberg, an open table format for large analytical datasets, with interoperability through Polaris Catalog and unified APIs.
Organizations can also deploy Cloudera engines like Apache Spark, Apache Kafka, or Trino, plus third-party or open-source solutions.
That matters. A common worry with new enterprise AI platforms is adding a second layer of technology dependence on top of existing cloud and data dependencies.
Cloudera says its open-standards architecture aims to limit that risk and keep its services from locking data into proprietary formats. The real interoperability will depend on the specific configurations, components, and services each organization adopts.
The platform also brings a self-service model through templates. Teams can deploy certain data, analytics, and AI services without hand-configuring the whole underlying infrastructure, while corporate governance policies are enforced on those deployments.
That mirrors the abstraction common in public cloud, now applied to on-premises setups that can stay physically inside the company.
AI agents start managing infrastructure too
A major feature linking Anywhere Cloud to the rise of agentic AI is its copilot.
Cloudera imagines users describing operations in natural language, with the system turning them into workflows over data or infrastructure management tasks.
That’s an interesting step for enterprise copilots. The first generation focused mainly on answering questions, generating code, or retrieving information through retrieval-augmented methods (RAG). Current agents also aim to take actions using external tools and systems.
That expands what they can do but raises the bar on permissions, observability, and governance. An agent that can query information carries a different risk than one allowed to change infrastructure or run operations on corporate systems.
Cloudera says Anywhere Cloud includes a centralized governance model based on zero trust, data traceability, and automatic policy enforcement. Those controls get more important as agents reach more enterprise tools.
The company is working with organizations like ADMIRAL Technologies, ExxonMobil, IQVIA, IXEN.ai, and Mastercard on Anywhere Cloud. That doesn’t necessarily mean full commercial deployments yet; Cloudera calls them Customer Design Partners, meaning they’re involved in shaping the platform’s design.
The launch also reflects a broader market shift. For years, much of enterprise modernization revolved around migrating to public cloud. Now AI is encouraging a more distributed architecture.
Compute costs, digital sovereignty, privacy, existing data center investments, and the sheer volume of accumulated data make bringing AI to the data more practical than moving all the data to AI systems in certain cases.
Anywhere Cloud is Cloudera’s bid to be that middle layer.
Frequently Asked Questions
What is Cloudera Anywhere Cloud?
It’s a data and AI platform for deploying and managing services across different environments, including public cloud, sovereign infrastructure, and private data centers.
Do you need to move data to the cloud to use it?
Cloudera says certain workloads can run where the data lives, avoiding transfer or copying between infrastructures. The exact capabilities depend on each architecture and application.
Which technologies does Cloudera Anywhere Cloud support?
It emphasizes open standards and technologies like Apache Iceberg, and supports engines such as Spark, Kafka, and Trino, along with third-party solutions.
What does agentic AI add to the platform?
Cloudera adds a copilot that can interpret natural-language requests to automate specific data workflows and infrastructure tasks.
via: cloudera
