Denodo has released Denodo Platform 9.5, aimed at a problem stalling enterprise AI rollouts. AI agents can reach data but cannot consistently trust or understand it across a business.
The release adds two features of note. Metric views let organisations define core business metrics such as revenue, churn and customer value once, governed centrally. Dashboards, business intelligence tools and AI agents then work from the same numbers rather than recalculating them differently.
An expanded enterprise knowledge graph, delivered through the Denodo Data Marketplace, gives AI systems visibility into governance, lineage and business context around data, not just the data itself. Denodo describes this context layer as the increasing blocker for agentic AI deployments at scale.
Denodo Platform 9.5 expands its Data Marketplace with a new 360 graph and asset extensions, enabling organizations to define an enterprise knowledge graph with a broader set of data ecosystem assets and the relationships that connect them. Teams can define and manage related assets such as extract, transform, and load (ETL) processes, consuming applications, notebooks, business glossaries, data dictionaries, governance controls, data product contracts, data sharing agreements, AI skills, and other business or technical artifacts.
The update also enhances reasoning in the Denodo Assistant to streamline data-view development, and expands connectivity across the data and AI ecosystem.
This release also expands connectivity across the modern data and AI ecosystem. These enhancements include improved support for unstructured data, new connectivity with Databricks and Azure AI Search, support for vector search with indexes for more efficient semantic and document queries, enhanced support for Delta tables and Databricks environments, and greater efficiency when using the Denodo Lakehouse Accelerator with Iceberg.
"These new capabilities enable us to enrich, connect, and contextualise our data assets, turning them into accessible, governed, and interconnected data products," said Jose Carlos Bermejo, head of data and analytics at Air Europa, which used a public beta.