Enabling AI with Starburst Enterprise

During Starburst Launch Point, you heard firsthand how organizations are using Starburst to build the foundations of modern data. Starburst is very excited to announce the release of the 474-e LTS. In addition to announcing our major AI innovations, including the Starburst AI Agent and Starburst AI workflows, the team is excited to announce several additional features that accelerate AI adoption by enabling what matters most–the data underlying AI.

Collectively, this release includes significant improvements in the following areas: 

• The Starburst AI Agent
• Starburst AI Workflows
• Enhanced Icehouse architecture 
• Connectors and Performance Improvements 
• Starburst Data Catalog 

 

The Starburst AI Agent

Starburst is launching the Starburst AI Agent, an agentic AI interface that operates across multiple domains. With the Starburst AI Agent, enterprises can build and scale AI applications faster, with reliable performance, lower cost, and greater confidence in security, compliance, and control. Starburst Agentic AI enables you to perform the following tasks, with additional functionality coming soon. 

Create data products using AI

Starburst AI Agents streamline the creation and documentation of data products by combining catalog metadata with user input and LLM-assisted editing, providing faster and consistent results.

Query data sources using natural language 

Use data products to query data sources using SQL generated from natural language input. 

 

Starburst AI Workflows

Starburst AI Workflows are a new suite of capabilities designed to accelerate enterprise AI adoption. This means moving AI strategy from experimentation to production by making governed, proprietary data instantly usable for a variety of use cases. These features are fully integrated with Starburst Enterprise itself, enabling the seamless integration of AI workloads alongside your analytics. Starburst AI Workflows are designed to address the key challenges enterprises face when scaling AI initiatives, particularly in terms of access, usability, and control. 

Highlights of this feature are listed below. 

Starburst AI Search 

The Starburst AI Search feature transforms structured, semi-structured, and unstructured data into vector embeddings stored in Iceberg tables, allowing AI agents to access and leverage the data. 

Starburst AI SQL Functions 

The Starburst SQL Functions enable analysts to apply generative AI directly within SQL queries, making it easy to analyze and transform unstructured text using functions like classification, sentiment analysis, and translation.

Starburst AI Model Access Management 

The Starburst AI Model Access Management feature provides a framework for managing access to proprietary and third-party AI models within a robust data governance framework. 

Read a complete list of supported AI models here.

 

Enhanced Icehouse Architecture

With this release, Starburst has taken its Icehouse architecture and pushed it further into the AI era, building on the core foundation based on Apache Iceberg and Trino. To this end, Starburst is excited to announce that Iceberg Materialized Views, Iceberg Data Products, and Iceberg Data maintenance are now in public preview.

These three features work together to provide a seamless lakehouse experience, allowing more opportunities than ever to make Iceberg your default table format without the operational overhead. By reducing complexity, eliminating manual maintenance tasks, and building for AI with Iceberg data products, Starburst Enterprise customers can spend their time discovering new insights. 

 

Connectors and performance improvements 

The last quarter has seen tremendous work on both our connectors and performance improvements. We are excited to announce that this area has seen continued growth in this release. 

Amazon S3 Tables

Starburst now supports Amazon S3 Tables, enabling users to query and federate data from AWS’s new managed Apache Iceberg format alongside over 50 other sources. 

Unity Catalog interoperability 

Starburst Enterprise continues to enhance its interoperability with the Databricks Unity Catalog. The latest updates ensure that customer changes made in Starburst Enterprise are instantly available and automatically reflected in Unity Catalog. The integration is a critical step towards our goal of truly interoperable computing, minimizing the need for separate copies of metadata, saving storage costs, and simplifying data management.

Starburst ODBC Driver 

The Starburst in-house ODBC driver has been updated and now includes improved data transfer performance. The new ODBC driver also supports both OAuthentication and Tableau’s generic ODBC connector at this time, providing timely updates and a better customer experience. 

Overall performance

This Starburst Enterprise release also includes exciting performance enhancements. This means that customers can now enjoy faster queries, particularly when using large amounts of intermediate data. It also means better efficiency in resource-constrained environments and enhanced dynamic filtering. 

Dynamic catalogs now support Warp Speed

Our dynamic catalog now supports Warp Speed. Dynamic Catalog Management enables you to define and manage catalogs directly through SQL statements, eliminating the need for manual updates to catalog configuration files. Warp Speed is Starburst’s proprietary indexing and caching layer. By combining Warp Speed and Dynamic Catalog Management, users will now experience enhanced performance improvements when utilizing object storage connectors. 

 

Starburst Data Catalog 

Starburst Data Catalog is now available in Private Preview. This change will help enterprise customers innovate their catalog layer, ensuring that their analytics workloads are optimized and ready for any AI workflows coming in the future. 

Replacing the Hive Metastore

Many enterprise customers rely daily on the Hive Metastore. Unfortunately, the technology underpinning Hive is not optimized for current analytics or AI workloads. The Starburst Data Catalog offers an alternative–a modern, enterprise-grade catalog designed for Iceberg, Trino, and beyond. 

Supports Iceberg 

The Starburst Data Catalog is designed to replace legacy Hive Metastore with a more secure, scalable, and interoperable foundation. Its design incorporates Apache Iceberg at its core, allowing you to manage schema and metadata for Iceberg tables without fragile workarounds. 

Enhancing data governance

This approach avoids data silos, and ensures strong data governance across your enterprise. With a single control plane for technical metadata, you get faster queries, cleaner governance, and better collaboration across teams. 

Please contact your account team to discuss the Starburst Data Catalog. 

 

Critical and breaking changes

This Starburst Enterprise release includes the following critical and breaking changes. Please view the full documentation for more information. Please view the full breaking change page for information. 

Java 11 minimum requirement

The minimum required Java runtime for the JDBC driver and CLI has been updated to Java 11. 

• If you are using an earlier version, you will need to upgrade to Java 11 or later.

Kinesis and Phoenix connectors

The Kinesis and Phoenix connectors have been removed due to their incompatibility with Java 24. 

• If you are using either of these connectors, please speak with our team to discuss the best alternative. 

Clickhouse connector changes

The ClickHouse connector now requires ClickHouse version 24.3 and Altinity version 22.3 as the minimum supported versions. 

• If you are using earlier versions of ClickHouse and Altinity, upgrade to these versions or later at your earliest convenience.

Custom Resource Definitions (CRDs)

The Custom Resource Definitions (CRDs) for the External Secrets Operator included in the Starburst Enterprise Helm Chart have been updated from v1beta1 to v1. 

• Upgrade your External Secrets Operator to v0.17, which fully supports v1.

Configuration changes 

Several configuration changes have also been implemented in this release. The changes include: 

• A number of dynamic filtering configurations have been replaced. 
• Some hash generation properties have been changed.
• HTTP client property prefixes have been renamed.
• Unity catalog metastore confirmation properties have been renamed in the Delta Lake connector. 
• Some Hive Metastore configurations using the Thrift API for Unity Catalog have been removed.

Please view the full documentation for more information.

 

Enhancing enterprise AI and analytics 

These capabilities reflect a simple idea: AI is only as good as the data that fuels it. Enterprise AI efforts will stall if the data used for agentic AI or LLM models is siloed, outdated, or inaccessible. 

Starburst is built to solve that problem—giving you open access, intelligent workflows, and robust governance across your entire data estate. Our platform supports your complete data journey from interactive analytics to autonomous AI agents.