Snowflake unveils new AI-powered data tools & expands platform
Snowflake has announced a suite of new platform updates and partnerships, expanding its offerings to make it easier for organisations to ingest, access, and govern data across their entire lifecycle and redefine the enterprise lakehouse for the artificial intelligence era.
The enhancements include new capabilities in the Horizon Catalog, the launch of Snowflake Openflow, and the introduction of Interactive Tables, Interactive Warehouses, and Snowflake Postgres. Each aims to support secure and effective data management, governance, and integration for businesses seeking to build agentic AI solutions and next-generation data-driven applications.
Horizon Catalog and open data access
Snowflake's Horizon Catalog now offers what it describes as a universal AI catalogue, enabling context and governance for AI across all data without a dependency on specific vendors. It allows organisations to integrate data from disparate sources, regardless of where it resides, spanning formats such as Apache Iceberg, Delta Lake, SQL Server, and Postgres.
"The enterprise lakehouse represents the evolution of how organisations manage and activate data for AI," said Christian Kleinerman, EVP of Product, Snowflake. "With advancements to Horizon Catalog, we're giving enterprises context and governance for AI across all their data by default - wherever it lives, and without vendor lock-in. Coupled with Openflow and Snowflake Postgres, it's now even easier for customers to connect and use their data securely, turning every dataset into the fuel for intelligence."
Snowflake has incorporated open application programming interfaces (APIs) from Apache Polaris and Apache Iceberg REST Catalog into Horizon Catalog. These integrations offer a centralised framework for governance, security, and access management in open table formats. External engines will soon be able to access data in Apache Iceberg tables through public previews, and manage this data with private previews.
To enhance resilience, business continuity and disaster recovery features for managed Iceberg tables are now available in public preview. The aim is to further protect critical enterprise data across the data lakehouse.
Interoperability
Supporting broader interoperability, Snowflake Openflow is now generally available, facilitating secure automation for data integration from nearly any source. This service is used by a range of enterprises to unify data and enable rapid deployment of AI-powered capabilities. Snowflake has also expanded its partnership with Oracle, allowing near real-time change data capture using Openflow for continuous streaming of updates into Snowflake's AI Data Cloud.
Peter Rogers, EVP, Head of Data and Technology, Americas at Merkle commented, "In the marketing industry, data is everything - and at Merkle, protecting that data and our customers' privacy is paramount. With Snowflake's Horizon Catalog, built from the ground up with security and governance at its core, we can ensure customer data, including sensitive identity information, is handled responsibly while safeguarding customer trust."
Near real-time experiences
The company is introducing Interactive Tables and Interactive Warehouses, intended to deliver low-latency, high concurrency analytics which enable teams to generate insights in sub-seconds. This shift to interactive analytics is designed to move beyond batch processing and provide immediate access to live data across business intelligence tools, supporting fast, intelligent applications and AI agents under one platform.
Alongside this, Snowflake has launched near real-time streaming analytics, currently in private preview. This technology enables organisations to act on incoming data streams within seconds and integrate live data with historical context for business use cases such as fraud detection, personalisation, and Internet of Things monitoring. Built-in support is provided for data sources such as Kafka and Kinesis.
Bringing workloads together
Following its recent acquisition of Crunchy Data, Snowflake has introduced Snowflake Postgres, a fully-managed service now available in public preview. It brings transactional, hybrid, and analytical workloads onto the Snowflake platform to address challenges traditionally caused by separating transactional data from analytical processing. The service is designed to allow enterprises to operate on both transactional and AI workloads natively within the same secure environment.
The open sourcing of the pg_lake extension further integrates Postgres with Snowflake's enterprise lakehouse. By using pg_lake, developers and engineers can query and manage Apache Iceberg tables directly via standard SQL within the familiar Postgres setting.
Additional advancements in Snowflake Unistore, powered by Hybrid Tables and now generally available on Microsoft Azure, enable building lightweight transactional applications on the platform. Enhanced security is being implemented with the introduction of Tri-Secret Secure support and periodic rekeying for Hybrid Tables, offering an additional layer of data protection that helps enterprises meet regulatory compliance requirements.
