Starburst unveils AI-ready platform for secure, governed data use

Starburst has introduced a set of new capabilities designed to enable seamless collaboration between human employees and AI agents across enterprise workflows.

The company's updated data platform now supports model-to-data architectures, multi-agent interoperability, and an open vector store on Iceberg, with the aim of allowing both humans and AI agents to reason, act, and decide more efficiently while maintaining oversight and governance.

Data access and governance

According to the company, the Starburst platform differs from legacy systems by enabling AI agents to securely access data wherever it exists-on-premises or in the cloud-without requiring data movement or relying on opaque retrieval processes. This federated, model-to-data approach, Starburst asserts, assists organisations in maintaining data sovereignty and compliance, particularly in highly regulated or cross-border contexts, while also reducing operational costs.

Starburst's new features include enhanced observability and visualisation tools for its agent framework. These enhancements allow organisations to monitor interactions with large language models (LLMs), set usage limits as guardrails, and review AI activities using dashboards. Starburst's agent is also capable of converting responses into charts and graphs, translating information into actionable visual insights for teams.

Justin Borgman, Chief Executive Officer and Co-Founder of Starburst, commented,

"With the Agentic Workforce, enterprises move beyond analyzing data to taking intelligent action. Our latest innovations bring models directly to governed data, allow AI agents to interoperate across multi-agent ecosystems, and provide open access to vector stores without lock-in. Starburst empowers organizations to scale AI securely and confidently across clouds, borders, and business-critical use cases."

Industry perspective

Andrew Brust, Founder and Chief Executive Officer of Blue Badge Insights, welcomed the development and emphasised the significance of governance and compliance. Brust stated,

"Enterprises have been looking for a way to bring structured data and governance into AI workflows, and the emergence of Model Context Protocol (MCP) has made that possible. But Starburst takes this much further. Not only does it unify vector store connectivity for RAG, but it provides AI agents with secure access to governed data products, leveraging Trino's federated architecture to do it all without data movement, thus ensuring compliance. The result is a pragmatic path for enterprises to scale AI with the trust, transparency, and regulatory control their business environments demand."

Platform features

Starburst's new platform aims to grant enterprises the flexibility to choose between model-to-data and data-to-model approaches when deploying AI. Among the new features are:

• Multi-agent ready infrastructure, including an MCP server and agent API, enabling the orchestration of multiple AI agents alongside the Starburst agent, so customers can build solutions for more complex tasks.
• Open and interoperable vector access, which offers unified access to vector stores for retrieval-augmented generation and search tasks across systems such as Iceberg, PostgreSQL with PGVector, and Elasticsearch. This enables organisations to select the best vector solution for individual workloads without vendor lock-in.
• Model usage monitoring and control, with monitoring and governance tools that help track, audit, and manage AI usage across different agents and workloads, aimed at controlling costs and ensuring regulatory compliance.
• Deeper insights and visualisation, with an updated conversational analytics agent capable of answering questions across various data domains using natural language, providing both narrative and visual responses, and discovering the right data for each query before processing.

Compliance and regulation

The company highlights the shift from dashboard-driven to autonomous, context-aware AI workflows, expressing that this transition places significant importance on answering regulatory and governance concerns-particularly for sectors like finance, telecommunications, manufacturing, and public services.

Matt Fuller, Vice President of AI/ML Products at Starburst, explained,

"Starburst's federated approach eliminates the need to centralize data while delivering consistent policy enforcement and transparent lineage. This means companies operating across European borders can confidently build AI and agentic workflows without compromising compliance. Our AI-ready lakehouse is designed with privacy, trust and performance at its core, giving teams governed access to the data that matters, whether they're training LLMs, deploying retrieval-augmented generation, or orchestrating multi-agent workflows, without limitations of legacy architectures or vendor lock-in."

Global deployment and policy enforcement

Starburst's platform is designed for use by enterprises operating in regulated regions, particularly the European Union. By offering federated access to distributed data, the platform allows organisations to query and analyse information where it resides and supports policy enforcement driven by metadata. This directly addresses compliance mandates such as GDPR and Schrems II, enabling companies to maintain governance wherever data is stored.

The latest capabilities are set to become generally available in the final quarter of the year.