Starburst offers enterprise and cloud solutions that simplify data querying across diverse formats, emphasizing flexibility and cost-efficiency. In this video, Justin Borgman, Co-Founder and CEO at Starburst, discusses the challenges of proprietary data formats and how open source formats are providing organizations with “optionality”. He goes on to talk about Starburst’s offerings and their focuses for the future. He says, “Building an infrastructure that’s going to give you optionality allows you to really manage cost performance, keeping costs as low as you can.”

What are the challenges of proprietary data formats and how are open formats helping?

• Borgman explains that his co-founders created Trino (formerly PrestoSQL) at Facebook, a query engine now used by major tech companies for data analysis. Starburst, the commercial entity behind Trino, offers enterprise and cloud editions to facilitate its use.
• Borgman highlights the challenges of proprietary data formats and vendor-specific SQL syntax, emphasizing the importance of data ownership for flexibility in querying and analysis.
• Borgman discusses how open formats like Apache Parquet and newer technologies like Apache Iceberg allow companies to control their data, avoiding vendor lock-in and enabling scalable, cost-effective operations.

How open source technologies benefit the community and an overview of Starburst’s products

• Borgman explains that open source technologies provide “optionality,” enabling customers to run systems themselves or use managed solutions like Starburst while retaining the freedom to switch back without major migrations.
• Open source also benefits from a supportive community that enhances the technology, exemplified by Trino’s geospatial functions developed by ride-sharing companies.
• Starburst products are Starburst Enterprise, a self-managed solution, and Galaxy, a cloud-hosted service that simplifies infrastructure management. Borgman emphasizes their cost efficiency compared to traditional cloud data warehouses.

The evolution of the Icehouse concept 

• Borgman discusses the evolution of the “Icehouse” concept within community-driven innovation for Lake Houses, highlighting Apache Trino and Apache Iceberg as components gaining industry traction as an open data stack.
• Borgman emphasizes Trino as the leading open SQL engine and Iceberg as the dominant open format, solidifying their positions through contributions at events like Trino Fest, supported by major players including Apple, LinkedIn, Snowflake, and Databricks.
• Borgman attributes the rise of Icehouse architecture to economic concerns over escalating Cloud Data Warehouse costs and the demand for long-term, flexible solutions.
• Borgman explains how the combination of Apache Iceberg and Apache Trino underscores their compatibility and enduring value in modern data architectures facilitated by Starburst’s efforts to simplify Icehouse implementation.
• Icehouse architecture, which separates storage (Apache Iceberg) and compute (Apache Trino), addresses the challenges of interoperability by allowing components like Spark or Delta to interact with either, providing more flexibility.
• The Icehouse architecture is suitable for a broad range of data processing and warehousing use cases, such as Business Intelligence (BI) and reporting.

How Starburst is leveraging Generative AI and the company’s current focuses

• Borgman explains how the Icehouse architecture enables organizations to efficiently analyze large volumes of data, deriving insights at a lower cost.
• Starburst leverages generative AI (GenAI) within the Icehouse architecture, particularly for natural language to SQL translation.
• Borgman recommends organizations prioritize data management and storage capabilities to support AI workloads efficiently and cost-effectively, given their resource-intensive nature.
• Borgman underscores Starburst’s deep commitment to the Icehouse architecture, emphasizing their leadership in Trino development and recent efforts to bolster Iceberg’s community.
• Starburst’s upcoming focus is on simplifying usability to provide customers with an intuitive, end-to-end experience akin to traditional Cloud Data Warehouses, yet leveraging the advantages of an open architecture.

Guest: Justin Borgman (LinkedIn)
Company: Starburst (Twitter)
Show: Let’s Talk

This summary was written by Emily Nicholls.