MinIO Unlocks On-Prem Enterprise AI Data for Databricks Customers

Even as enterprises increasingly standardize on Databricks for advanced analytics and AI, a growing share of their most valuable data remains on-premises due to scale, performance, cost, and data sovereignty requirements. Historically, making that data available to Databricks required complex pipelines, duplicate datasets, and separate governance layers, resulting in delayed time-to-insight, increased costs, operational risk, and ongoing overhead.

AIStor Table Sharing addresses that friction by embedding Delta Sharing directly into the object store. Delta Sharing is Databricks’ open source approach that enables customers to share live data across platforms, clouds, and regions with strong security and governance. Table Sharing enables federated, enterprise-scale analytics without data movement, copies, or lock-in.

“Enterprises shouldn’t have to move massive datasets just to analyze them,” said AB Periasamy, co-founder and co-CEO, MinIO. “Today, all data is AI data, and as AI blurs the lines between on-premises and cloud environments, data gravity remains a hard reality. Powered by Delta Sharing, AIStor Table Sharing removes that constraint by allowing data to be accessed and shared where it lives, providing faster insights, lower risk, and simpler operations as enterprises scale AI across hybrid environments.”

“Customers consistently ask us to be able to govern and share data stored in and out of the cloud. Our partnership with MinIO is a testament to the power of an open data ecosystem,” said Stephen Orban, SVP, product ecosystem and partnerships, Databricks. “By natively integrating Delta Sharing, MinIO enables enterprises to securely connect their on-premises data to Databricks without complex replication, accelerating time-to-insight for hybrid workloads.”

Unifying Enterprise Data for AI at Scale
AIStor Table Sharing is built on AIStor Tables, the Iceberg V3-native foundation for modern data lakehouses running at enterprise scale on-premises and in hybrid environments. AIStor Tables combines MinIO’s high-performance, S3-compatible object storage with integrated Iceberg table catalogs, metadata, REST API, and open sharing standards, allowing table shares to be defined, governed, and published directly from the same system where the data is stored.

By bringing structured and unstructured data together in a single platform, AIStor Tables transforms AIStor into a true AI data store, feeding analytics engines and GPUs directly without data duplication, architectural sprawl, or operational overhead. AIStor Table Sharing extends this foundation, enabling Databricks customers to unify access to their on-premises data for analytics and AI while preserving performance, cost efficiency, and control.

As AI raises the stakes for speed, scale, and governance, AIStor Table Sharing redefines how enterprises can unify and connect their cloud compute with their on-premises data stores:

Open Sharing, Built In: Native implementation of the Delta Sharing enables seamless, standards-based data sharing without proprietary lock-in
Instantly Analyze Data Where It Lives: Access live on-premises data without replication, eliminating costly data movement and copies while preserving performance and control
Multi-Table Format Flexibility: Works with both Delta and Apache Iceberg tables, giving enterprises the freedom to standardize on open formats as analytics and AI needs evolve
Native Databricks Integration: Integrates directly with Databricks via the Delta Sharing protocol, enabling enterprises to extend their cloud analytics compute to on-premises data without added delays, costs, or complexity
Built for Real-World Architectures: Designed for on-premises, hybrid, and multi-cloud environments, aligning with how enterprises actually store and govern data today

Purpose-Built for Enterprise Data Environments
MinIO and Databricks share a growing base of large enterprise customers across manufacturing, financial services, energy, retail, and logistics. These organizations generate and retain massive volumes of on-premises data for valid operational, economic, and regulatory reasons and want to apply Databricks’ analytics and AI capabilities to that data in place.

Flexibility at Enterprise Scale
AIStor Table Sharing reflects how enterprises actually operate today: multi-format, hybrid, and at massive scale. By supporting both Delta and Apache Iceberg tables, MinIO avoids locking customers into a single analytics path or proprietary sharing model.

All MinIO AIStor editions share the same binary and core architecture, differing only by licensed features, supported scale, and support entitlements. This allows customers to adopt AIStor table sharing today and seamlessly scale as requirements grow.

Availability
AIStor Table Sharing is now generally available with MinIO AIStor.