Snowflake He has brought all his novelties to his annual user conference Snowflake Summit 2025focused above all on facilitating companies management, analysis and use of your data within the framework of AI. In addition to taking advantage of the event to confirm the purchase of Crunchy Data.
The novelties cover data engineering, computer yield, analytics and agricultural capabilities. One of them is snowflake Openflowa multimodal data intake service, which allows users to connect with almost any source of information, regardless of their architecture.
Available in general in AWS, OpenFlow has been designed to deal with data fragmentation and to avoid manual processes, joining different types and data formats to be able to display innovations based on AI.
It is based on open standards, which allows to centralize data integration into a unified platform, in addition to allowing the centralization of data integration into a single platform, avoiding dependence on a single supplier and having interoperability of guaranteed architecture.
Promoted by the Apache Software Foundation Apache Nifi project, OpenFlow facilitates the creation of personalized connectors, which are executed without barriers on the Snowflake managed platform.
In addition, it allows users to take advantage of their data throughout the entire life cycle, adapting to data standards and different business needs.
It already has hundreds of preconfigured connectors and processors simplify and accelerate data integration from various sources to different destinations, including storage of cloud objects and messaging platforms.
Among these sources are Box, Google Ads, Microsoft Dataaven, Microsoft SharePoint, Oracle, ProofPoint, Servicenow, Workday and Zendesk.
Snowflake news in computation for data storage
Snowflake has also led to its event its computing innovations, which improve in performance, and usability, as well as in Price-relevance relationship.
Among them is Standard Warehouse – Generation 2 (GEN2), an updated version of Standard Warehouse Virtual, with latest generation hardware and other novelties that accelerate the data of data analysis, and achieve a faster performance than Managed Spark.
Adaptive Computeanother novelty that is now in the test version, is a service that facilitates the management of resources thanks to automatic optimization of the size and shared use of said resources.
Data storage generated with Adaptive Compute, known as adaptive warehouses, improve users.
Snowflake Intelligence y Data Science Agent
Snowflake Intelligencewhich will be available soon in trial version, allows technical and non -technical users to ask natural language questions, and get value data from structured tables and unstructured documents.
Intelligence is driven by the LLM of the latest generation of Anthropic and Openai, which are executed within the Snowflake safe perimeter, and is driven internally by Cortex Agents (test version). All through an interface without code.
As to Data Science Agentwhich will be available in the coming weeks in a private test phase, is a digital assistant based on agents, designed for data scientists, which allows automating routine tasks for the development of Machine Learning models.
Data Science Agent uses Claude from Anthropic to divide the problems associated with Machine Learning’s workflows in several steps. Among these steps are data analysis, data preparation, functions engineering and training.
Cortex AISQL, SnowConvert AI y Marketplace
Snowflake has also announced several novelties that extend Cortex AI. Are SnowConvert AI y Cortex AISQL. The first is an agricultural automation solution that accelerates migrations to snowflake from inherited platforms. It allows data professionals to modernize their infrastructure more quickly and with less effort.
When the data comes to Snowflake, Cortex Aisql incorporates the generative AI directly in customer consultation engines. Thus, equipment can get multimodal data information, and create flexible flows using SQL.
In addition, Snowflake has announced new agents products in Marketplacefocused on accelerating the adoption of agricultural AI in companies. Among them, Cortex Knowledge Extensions, available soon, that allows companies to enrich their applications and AI agents with unstoping data patented from external suppliers.
At the same time, it allows suppliers to protect their intellectual property and guarantee adequate attribution. Users can also access a selection of articles and business content of The Associated Press, which will help them improve the usefulness of the results in their AI systems.
The company has also presented the Semantic Models exchangenow in a private test version, which allows users to integrate structured data prepared for AI in their applications and agents of Snowflake Cortex AI. Both internal equipment and external suppliers, such as carto, CB Insights Ipinfo and Truestar.
