The ability to access and analyze data from multiple data sources has always been critical to business intelligence (BI), driving the evolution of the market for data integration products to ensure data from operational workloads is available for analytics use cases. Despite that, data silos continue to be a perennial challenge with the potential to stifle innovation given the speed and complexity of data processing required to support automated agentic decision-making. Data platform providers have been expanding investment in data federation and zero-ETL processes to facilitate access to data across multiple data platforms. Providers with a strong track record in query acceleration and federation, such as Starburst, are well-placed to benefit from evolving requirements for real-time querying of distributed data.
Starburst was founded in 2017 to build a commercial business around the Trino distributed SQL query engine. Trino was initially used to perform SQL-based data analysis in Apache Hadoop and object storage environments but is now more commonly used to accelerate the analysis of data stored in open table formats, such as Apache Iceberg. The combination of Trino and Iceberg forms the basis of the Icehouse architecture at the heart of the company’s Starburst Enterprise Intelligence Platform. In addition to the core connectivity and analytics engine functionality, the Starburst Enterprise Intelligence Platform also provides capabilities for creating and managing data products, as well as agent-driven conversational analytics. The company was rated as a Provider of Merit in the 2026 ISG Buyers Guide for Analytic Data Platforms as well as the 2026 ISG Buyers Guide for Data Products. Starburst announced in early 2026 that it surpassed $100 million in annual recurring revenue the previous year. The company has raised more than $400 million in venture funding, including a $250 million Series D funding round at a $3.35 billion valuation in early 2022 and a more recent strategic investment from Citi.
Many providers are repositioning their products in relation to providing context for artificial intelligence (AI) and agents. Multiple layers of context are required, and no single product can currently address all requirements, but Starburst can claim to address more than many. The company’s Starburst Enterprise Intelligence Platform architecture was launched in May this year and provides a multi-layered approach to enterprise intelligence delivered via the Starburst Galaxy managed service or Starburst Enterprise for self-managed on-premises, cloud and hybrid deployment. The foundation of the Enterprise Intelligence Platform is provided by the combination of Trino and Iceberg-based data lakes, with Starburst providing the ability to ingest, persist and analyze data in the Iceberg format, as well as automating the management of Iceberg tables with functionality addressing data compaction, purging and table optimization. Interoperability between multiple data platforms with support for Iceberg is enabled by Starburst’s Iceberg Connector, while Trino’s distributed SQL query capabilities enable queries to be federated across multiple source platforms, avoiding the need to ingest and integrate the data prior to its analysis.
This data federation capability offers the potential to overcome the perennial challenge of data silos. Two-fifths (40%) of participants in ISG’s Data and AI Market Lens Study cited data silos and fragmentation as a primary data challenge. Additionally, the separation of the
In addition to this analytics engine layer, the foundation provides an enterprise context layer through which users can develop, manage, govern and share data products based on agreed business definitions, business rules and metrics. Starburst also recently added a new Data Products as Code capability that enables users to create and share data products as YAML files, enabling version control and facilitating inspection of and interaction with data products by AI agents. The Starburst Enterprise Intelligence Platform also provides an agentic control plane layer to enable agent governance, privileges, observability and guardrail functionality.
While the Starburst Enterprise Intelligence Platform can be used to enable analysis using multiple analytics tools, applications and agents, it also includes Starburst’s own AI Data Assistant (AIDA), which was unveiled in April to accelerate decision-making using Starburst Galaxy and Starburst Enterprise. Starburst AIDA is a natural language analytics interface designed to enable business users and decision-makers to analyze data via Starburst without writing SQL queries, and it can also be used by data engineers for data exploration.
Using the underlying Starburst AI agent, AIDA translates natural language queries into SQL and returns results as summaries, as well as tables and visualizations. AIDA tailors the format of results via three default personas designed to match the requirements of
I have previously cited some potential challenges with conversational analytics products designed to enable anyone and everyone to generate insights from data. As a result of these challenges, I assert that by 2028, 4 in 10 enterprises will be experimenting with vibe analytics products to accelerate the creation of proof-of-concept analytics initiatives through rapid prototyping by data consumers. Starburst’s combination of personas and pre-defined guardrails is designed to overcome some of these challenges. These capabilities will be complemented by the forthcoming introduction of Context Studio, which Starburst has previewed as an environment for harvesting business definitions and metrics from BI tools, catalogs and integration pipelines, as well as query history to automate the creation of context. In the interim, I recommend that enterprises evaluating products to facilitate the modernization of their data estate to support agentic decision-making include Starburst in their assessments.
Regards,
Matt Aslett