What’s New in Snowflake Summit 2025
The main innovations and improvements revealed at the Snowflake summit 2025 are highlighted in this blog, but there is more to come.
1. Get Started with Agentic AI with No Code
Agentic AI refers to AI systems that proactively assist users by understanding context, initiating actions, and automating tasks. Snowflake’s agentic AI experiences enable both technical and non-technical users to utilise and interact with AI and data using natural language, without writing a single line of code. The key features highlighted are:
• Data Science Agent (coming soon in private preview): Your ML copilot that can automate feature engineering, training, and more using plain English.
• Cortex AISQL (public preview): Cortex AISQL enables users to utilize SQL to analyze documents, images, and other unstructured data types. Now includes schema-aware table extraction from complex PDFs.
• Cortex AI Agents (public preview): You can use natural language to ask questions, analyze structured and unstructured data, and take action, all within the Snowflake platform.
Bonus: All of this operates within Snowflake’s secure perimeter, maintaining governance and privacy by default.
2. Unified Data Engineering and Interoperability
Snowflake is solving rigid data pipelines and slow ingestion issues with new tools built for flexibility and speed. Let’s have a look:
• Snowflake OpenFlow (powered by Apache NiFi): A managed, extensible, multimodal low-code ingestion platform. It includes hundreds of connectors to eliminate data silos. To facilitate smooth ETL processing for AI, OpenFlow transforms data movement within Snowflake.
• Native dbt Integration (coming soon): With Git and AI Copilot built-in, you can develop and run dbt models natively within Snowsight UI.
• Iceberg Table Enhancements: You can work effortlessly with external catalogs using the Iceberg REST API. It supports VARIANT data types and Merge-on-Read for semi-structured data.
• DevOps-Ready: Snowflake Workspaces now support custom Git URLs and Python 3.9. Terraform provider is now generally available.
3. Analytics with AI-Driven Acceleration
Analytics just got smarter, faster, and easier to scale with the following updates:
• Snowflake Semantic Views (public preview): It provides excellent support for BI teams, and you can create consistent, reusable business metrics and entities within Snowflake.
• Gen2 Warehouses: Generation 2 (Gen2) is a modernized version of Snowflake’s standard warehouse, featuring enhanced hardware and improved performance capabilities. It provides a 2.1x performance boost for analytical queries with enhanced scans, DELETEs, and MERGEs.
• SnowConvert AI: A free, automatic tool for auto-converting legacy ETL/BI code to Snowflake. It reduces risk and saves migration costs.
4. Smarter Marketplace to Share, Sell & Collaborate Easily
Snowflake is making it easier to monetize AI apps and datasets using:
• Cortex Knowledge Extensions: While maintaining IP protection, embed real-time info from Stack Overflow, AP, Packt, and more into your apps.
• Semantic Model Sharing (private preview): Rely on natural language to explore datasets across internal teams or external partners.
• Agentic Snowflake Native Apps: You can easily build and sell no-code, AI-powered apps directly through the Snowflake Marketplace.
5. The Future of Platform: Adaptive, Governed & Secure
Snowflake isn’t just innovating; it’s redefining what a modern data platform should be in the following ways:
• Adaptive Compute (Adaptive Warehouses): You can automatically adjust resources based on workload, delivering speed without overspending.
• Horizon Catalog (private preview): It enables users to easily discover and govern external data sources, such as dashboards and relational databases.
• Copilot for Governance (coming soon): Users can utilize this feature to manage metadata, access policies, and security using natural language.
• Security Upgrades: To improve account protection, enhanced MFA, password policies, and a revamped Trust Center are being implemented.
• Snowflake Trail: You can debug pipelines and optimize your generative AI agents and apps using new observability features.
Summing Up
Snowflake Summit 2025 showcased the company’s push toward an Agentic AI enterprise data platform. Snowflake’s advanced features, including Cortex AISQL, OpenFlow, dbt integration, and native app frameworks, are democratizing AI and making it easier for users. Whether you’re a data scientist, engineer, or business leader, Snowflake’s latest innovations promise faster time-to-insight, reduced complexity, and deeper control over your data universe.
Back
