We had an amazing time last week at the Databricks Summit. Collectively, our team attended over three dozen sessions ranging from new governance capabilities in Unity Catalog, to frameworks that streamline AI development, particularly RAG architectures and Lakehouse Federation. Plus, we spoke with over 500 data leaders at the Acante booth about how to make data access governance easy for data teams.
Thought we’d share our top take-aways from the show.
I. Governance of the Entire Data Estate is Hard
Security and privacy are under pressure - a top headline Databricks CEO, Ali Ghodsi, pointed out in his keynote. He highlighted how the data estate is increasingly fragmented. We couldn’t agree more. Customers are bringing in ever-increasing amounts of data from dozens of sources, transforming it, and sharing it. Streamlining access governance for this dynamic data from the standpoints of compliance & privacy, data security, and orchestration is key to unblocking data platform & AI initiatives.
II. Unity Catalog will soon support Attribute-based Access Controls adding long missing control primitives
Expressing fine-grained access controls using low-level GRANT statements and User-Defined Functions (UDFs) for column/row-level masking is painful, to say the least! By introducing ABAC primitives (in Private Preview) into the lexicon, Databricks has filled a feature gap vs. Snowflake allowing use of dynamic tags to define data assets. This opens the door to 10x fewer policies and global masking controls. However, managing these ABAC policies consistently on each metastore, and then across the multiple platforms in your data estate will remain a challenge. And keeping track of the effective access granted in one place will become more critical than ever.
III. Data Intelligence is your edge with AI
Data Intelligence is about leveraging your proprietary data in your AI applications. Databricks made a slew of announcements including an agent framework, a vector search offering, and RAG architectures that will make it easier for you to use your organization’s data to build your competitive AI edge.
At the same time, in his keynote, Ali revealed that 85% of GenAI projects have not made it to production and cited security and privacy as a top reason for that. Platforms that can address this need will become key to enterprise AI adoption.
Where Acante Comes In
Acante empowers data teams to overcome barriers to democratization of data access and accelerate their Data+AI projects. Acante has built a comprehensive multi-platform data & AI access governance solution for modern data stacks built around Databricks, Snowflake, S3, Redshift, ADLS, Kafka, Tableau, PowerBI, etc.
Acante centralizes the necessary data context to provide a simpler and faster way to discover, track, right-size, protect and provision access to their most sensitive data. Modern orchestration workflows and automated insights empower data teams to deliver frictionless access to data while meeting security and compliance mandates.
