Databricks News and Vouchers: Highlights from Data + AI Summit 2025
On June 12, 2025, the Data + AI Summit, Databricks' flagship annual event, concluded in San Francisco, gathering over 20,000 data and AI professionals from around the world. The event introduced a series of announcements and innovations set to transform the data, artificial intelligence, and cloud collaboration ecosystem. Below, I share a summary of the main news unveiled at the event, with brief descriptions for easy understanding.
1. Databricks Lakeflow: Unified Data Engineering
Databricks Lakeflow was launched as a comprehensive solution for data ingestion, transformation, and orchestration, integrating managed connectors for enterprise applications, databases, and data warehouses. A highlight is Zerobus, an API enabling real-time event data ingestion with high throughput and low latency, making large-scale data usage for analytics and AI easier.
2. Unity Catalog: Intelligent Governance and Automation
Unity Catalog received new features to unify data and AI governance across different formats, clouds, and teams. Notable updates include:
- Attribute-Based Access Control (ABAC): Enables flexible access policies using tags, now in beta for AWS, Azure, and GCP.
- Tag Policies: Ensure consistency and security in data classification and usage across the platform, also in beta on major clouds.
3. Data Sharing and Collaboration
Improvements were announced to facilitate secure data sharing between organizations, including “clean rooms” that allow collaboration without compromising data privacy or security.
4. Full Support for Apache Iceberg™
Databricks now offers full support for Apache Iceberg™, expanding open-format data management possibilities and making integration with various tools and platforms easier.
5. Spark Declarative Pipelines
The platform introduced Spark Declarative Pipelines, an evolution for developing data pipelines in a declarative, scalable, and open way, boosting productivity and standardization for data engineering teams.
6. Databricks SQL and Free Edition
General availability of Databricks SQL was announced, along with a new free edition of the platform, democratizing access to advanced data analytics and intelligence resources for organizations of all sizes.
7. MLflow 3.0: AI Observability and Governance
MLflow 3.0 arrives with improvements for experimentation, observability, and governance of AI models, streamlining the complete machine learning project lifecycle within the Databricks ecosystem.
8. Mosaic AI and Agent Bricks
Mosaic AI introduced new features for developing intelligent agents, including Agent Bricks, which enables the creation of self-optimizing agents using proprietary company data, accelerating the practical adoption of generative AI and autonomous agents.
9. Lakebase: Public Preview
The Lakebase concept was presented in public preview, offering an innovative approach for managing transactional and analytical data in a single environment, simplifying operations and accelerating insights.
10. Power Platform Connector
The new Azure Databricks connector for Power Platform enables real-time, governed data access for Power Apps, Power Automate, and Copilot Studio, expanding integration possibilities between data platforms and productivity tools.
These innovations reinforce Databricks' commitment to leading in data and AI, offering increasingly integrated, secure, and accessible solutions for organizations across all sectors. Stay tuned, as these updates are sure to impact the market in the coming months.
Top comments (0)