Cloudera is expanding its services to include running AI inference capabilities and its Trino-powered data warehouse directly within customer data centers. The company is also enhancing its Cloudera Data Visualization tool with modern AI-driven features designed to unify workflows across multi-cloud, edge, and on-premises environments, according to a company release.
The move comes as organizations increasingly transition from AI experimentation to production deployments, with a growing focus on secure and governed data access regardless of location – shifting the conversation from where data is stored to how it can be reliably utilized. This trend underscores the importance of data governance and security as AI becomes more integrated into business operations.
On-Premises AI Inference Reduces Sensitive Data Transfer
A recent Cloudera report, “The State of Enterprise AI and Data Engineering,” found that nearly half of companies retain their data within private data warehouses. Enabling AI applications to access this data within secure environments can reduce the need to transfer sensitive information outside the organization, impacting security, compliance, and operational risks, the company stated.
Leveraging NVIDIA Technologies for Local AI Model Deployment
Cloudera’s on-premises AI inference service will be powered by NVIDIA technologies, allowing organizations to deploy and scale a variety of AI models within their data centers. This includes support for open-source NVIDIA NeMo models, as well as use cases like large language models (LLMs), fraud detection, computer vision, and speech technologies.
The service is built on the NVIDIA AI ecosystem, utilizing NVIDIA Blackwell GPUs, the NVIDIA Dynamo-Triton inference server, and NVIDIA NIM microservices for model serving. According to Cloudera, this approach aims to deliver governed AI deployment with more predictable costs compared to variable cloud models, although also giving organizations greater control over latency, compliance, and data privacy as projects move into production.
“Cloudera Data Warehouse with Trino” Now Available On-Premises
The “Cloudera Data Warehouse with Trino” is now available in customer data center environments, offering centralized security, comprehensive governance, and observability across data assets, designed to accelerate access to analytical insights. The integration of AI-powered analytics and visualization tools aims to help organizations transform data into actionable results while maintaining security, compliance, and operational control.
“Cloudera Data Visualization” Updates Enhance Discoverability and Simplify Management
Cloudera has announced improvements to its “Cloudera Data Visualization” tool, designed to deepen insights and simplify AI-powered workflows both within and outside of data centers. These enhancements include:
AI-powered explanations: To generate instant summaries and contextual insights for charts and graphs without manual writing.
Flexible AI features: To handle transient issues and provide detailed usage analytics for monitoring and optimization.
AI query tracking and logging: By logging each query with a message ID, timestamp, and question to ensure traceability.
Simplified administrator management: Through easy assignment of administrative roles using updated configuration criteria to streamline single sign-on (SSO)-based setup, removing embedded credentials and manual user upgrades.
What Cloudera and NVIDIA are Saying About the Expansion
“These developments give our customers an unparalleled level of control and flexibility,” said Leo Broniek, Chief Product Officer at Cloudera. “With both our AI inference service, data warehouse with Trino, and data visualization tools available within data centers, organizations can deploy AI and analytics securely where their most critical data resides. This empowers them to drive innovation and extract insights without compromising data security, compliance, or operational efficiency.”
“The true value of enterprise data is realized when AI can be deployed flexibly and securely in the same location as that data,” said Pat Lee, Vice President of Strategic Enterprise Partnerships at NVIDIA. “Our collaboration with Cloudera enables customers to deploy and scale AI inference using NVIDIA Blackwell GPUs, Dynamo-Triton technologies, and NIM microservices, delivering full control, predictable economics, and high efficiency for data centers.”
Session at Developer Week in February 2026
Cloudera will participate in Developer Week from February 18-20, 2026, and plans to host a session on building a cloud-native open lakehouse using Apache Iceberg.
Cloudera is a company focused on data, analytics, and AI platforms for the enterprise, delivering services across multi-cloud, edge, and data center environments.
