A forward-looking conversation on architecting data for AI where speed, scale, and governance converge. Insightful perspectives on building resilient platforms in a landscape shaped by continuous disruption.
To start, Justin, can you share a bit about your journey as a software engineer and how those early experiences influenced your path to founding Starburst?
My background as a software engineer gave me firsthand insight into the complexities of building technology that solves real-world problems. Early in my career, I founded Hadapt, which introduced “SQL on Hadoop,” transforming Hadoop into a tool for analytics and laying the groundwork for today’s Lakehouse architecture. After Hadapt’s acquisition by Teradata, I witnessed the growing inefficiencies of traditional data warehouses as businesses grappled with massive data growth and cloud adoption. Discovering Trino, an open-source distributed SQL engine, inspired me to co-found Starburst with a mission to enable efficient, scalable data analytics without the constraints of proprietary systems.
Starburst is at the forefront of lakehouse technology. What inspired the creation of this platform, and how do you see it transforming the way businesses approach AI-driven insights?
Starburst was inspired by the need for a better way to analyze data spread across multiple systems without relying on traditional warehouses. As businesses faced the rapid rise of big data, cloud computing, and AI, the inefficiencies of centralized, proprietary data storage became clear. Trino, a distributed SQL engine originally built for Facebook, showed a new path forward: querying data where it lives. This became the foundation for Starburst.
By eliminating the need for duplicating or migrating data, Starburst allows organizations to access insights faster and at a lower cost. This transformation accelerates AI-driven innovation, as businesses can leverage all their data seamlessly for machine learning, advanced analytics, and real-time decision-making.
AI workflows are often challenging to operationalize. How does Starburst’s integration of vectors and data streamline this process, particularly with large language models (LLMs)?
Operationalizing AI workflows, especially for large language models (LLMs), often entails overcoming hurdles related to data access and transformation. Starburst’s integration of vectors with its distributed SQL engine makes this process significantly more efficient. By enabling organizations to query, combine, and analyze both structured and unstructured data where it resides, Starburst eliminates the need for cumbersome data migration and preprocessing.
For LLMs, which thrive on rich, context-aware datasets, Starburst’s approach delivers a unique advantage. The ability to integrate and process diverse data sources in real-time not only accelerates feature engineering but also improves model performance by providing a richer, more contextual data foundation. This holistic platform helps to streamline model training and inference, speeding up development cycles and ensuring that AI workflows are operationalized more quickly and with higher accuracy. The result is a scalable, cost-effective solution that enhances the use of data at every stage of AI deployment, empowering businesses to generate insights and take action faster than ever before.
Cost and efficiency are critical in enterprise AI. Could you elaborate on how Starburst’s technology enables businesses to reduce AI compute costs by up to 50% without compromising on performance or scalability?
Starburst reduces AI compute costs by up to 30 – 50% by streamlining data processing and eliminating the need for data duplication and movement, which are major drivers of infrastructure expenses. Our platform allows businesses to query, transform, and analyze data directly where it resides, significantly cutting costs without sacrificing performance or scalability.
Security and compliance are paramount in AI-driven operations. How does Starburst ensure secure and governed interactions between enterprise data and AI models?
Starburst ensures secure and governed interactions between enterprise data and AI models through robust governance and access controls. Our platform enforces compliance with industry regulations like GDPR, providing full visibility into data usage and access. By enabling role-based permissions, Starburst safeguards sensitive information while ensuring only authorized users can interact with critical datasets.
Adapting to AI’s rapid evolution is a pressing concern for businesses. How does Starburst’s data platform for AI help organizations future-proof their AI investments and prepare for emerging trends?
Starburst’s data platform for data apps and AI future-proofs AI investments by enabling businesses to scale seamlessly as AI evolves. Our ability to query data across multiple sources without moving it allows organizations to integrate emerging technologies and trends like AI and machine learning without costly infrastructure changes. By combining the flexibility of a data lake with the performance of a warehouse, Starburst ensures businesses can easily adapt to increasing data complexity and growing AI demands. This architecture supports real-time analytics, enabling enterprises to stay agile, scale efficiently, and leverage data for new AI-driven innovations as they emerge.
Open standards and interoperability are becoming key factors in making AI investments resilient. What role do they play in Starburst’s platform, and why are they essential in today’s data landscape?
Open standards and interoperability are central to Starburst’s platform, particularly through our use of Apache Iceberg and Trino in the Icehouse architecture. Icehouse combines the best features of data lakes and warehouses, leveraging open formats and SQL, ensuring that data is portable and customers are not locked into proprietary systems. This open approach gives organizations full control over their data and queries, enabling them to seamlessly scale and adapt as their needs evolve.
Trino’s ability to query diverse data sources and Apache Iceberg’s efficient table management are essential in today’s data landscape, where businesses require flexibility, performance, and scalability across multi-cloud, hybrid, and on-premises environments. With Starburst’s Icehouse, we empower customers to manage and analyze large datasets across any infrastructure, all while maintaining the freedom to use open technologies and avoid vendor lock-in, making it an ideal solution for future-proofing AI investments.
Bridging the gap between data scientists and business stakeholders can often be a challenge. How does Starburst empower non-technical users to access and leverage AI insights effectively?
Starburst empowers non-technical users by simplifying data access and enabling intuitive SQL queries across disparate data sources. Our platform allows users to analyze large datasets without needing deep technical expertise, providing an easy interface to access insights from AI models.
Starburst ensures fast, scalable analytics, enabling business stakeholders to make data-driven decisions. With seamless integration into existing business workflows, Starburst breaks down silos, allowing non-technical users to interact with complex data and AI insights effectively, fostering collaboration between data scientists and business teams.
Looking ahead, how do you envision the relationship between data platforms and AI evolving in the next decade? What major trends do you anticipate driving this change?
AI is a space that is ripe for continual disruption over the next decade. No winners or losers have been declared, so every player must be ready to respond and poised to adapt as market dynamics change. We’ve already begun to see disruption at every layer of the AI stack including energy, infrastructure, data layer, and apps and agent layer.
Finally, as democratized access to AI capabilities becomes more prevalent, what broader implications do you see for businesses and society, and how is Starburst preparing to lead in this transformation?
We are working with companies of all sizes and around the globe to leverage the power of AI. In some cases, leading edge startups are directly embedding Starburst into their customer-facing applications because we offer a speed, governance and economic model where that makes sense. For some of our largest companies we are helping them in their journey to AI with starting on specific pilots and proof of concept before they scale across the enterprise and around the globe. At the end of the day, we’re meeting our customers where they are as they ready their business to capitalize on AI.
A Quote or Advice from Justin
“To truly unlock the transformative potential of AI, businesses must eliminate data silos and enable seamless access to both structured and unstructured data, in a secured and governed manner. Starburst’s data platform for AI empowers organizations to accelerate the development and deployment of data apps and AI and do so more cost effectively. This enables faster innovation, lowers costs, and allows businesses to scale their AI efforts as technology evolves, all without the need for costly infrastructure changes. The future of AI is built on the foundation of an open, interoperable data landscape.”
Justin Borgman
Co-founder and CEO at Starburst
Justin Borgman is a subject matter expert on all things big data & analytics. Prior to founding Starburst, he was Vice President & GM at Teradata (NYSE: TDC), where he was responsible for the company’s portfolio of Hadoop products. Justin joined Teradata in 2014 via the acquisition of his company Hadapt where he was co-founder and CEO. Hadapt created “SQL on Hadoop” turning Hadoop from a file system to an analytic database accessible by any BI tool. He founded Starburst in 2017, seeking to give analysts the freedom to analyze diverse data sets wherever their location, without compromising on performance.