Snowflake Summit 2025 Recap: 20+ Next‑Gen Feature Rollouts

Snowflake Summit 2025 took place in San Francisco from June 2–5, delivering a clear message: data quality is the foundation of AI. Snowflake CEO Sridhar Ramaswamy kicked things off by stating, "There’s no AI strategy without a data strategy". He emphasized the need for simplicity and trust when it comes to data. NYSE President Lynn Martin echoed this sentiment, saying that the "sanctity" of data is essential for enterprise AI to work. One of the biggest surprises was Snowflake’s acquisition of Crunchy Data. Here’s a full breakdown of everything released, announced, or presented from June 2 to June 5, 2025.

Day 1 — Monday June 2, 2025: Opening Keynote & Sneak Peeks

Snowflake’s CEO, Sridhar Ramaswamy, opened the Summit with a casual fireside-chat style keynote featuring OpenAI’s CEO, Sam Altman, moderated by investor Sarah Guo. The conversation set a big-picture tone about AI and data.

Sam Altman’s appearance was more than just symbolic. His 15-minute segment revealed candid thoughts about AI’s current limitations and future potential. Altman compared today’s AI systems to "a company intern, a junior employee at best", but predicted that within one to two years, we would see agents capable of discovering new knowledge rather than just handling repetitive cognitive tasks. When asked about enterprise AI adoption, Altman’s advice was blunt: "Just do it. There’s still a lot of hesitancy... But as a general principle of technology, when things are changing quickly, the companies that have the quickest iteration speed and make the cost of making mistakes the lowest win".

Day 1 featured no major product launches. Instead, it focused on vision and context. The biggest surprise wasn’t a new feature but an acquisition: Snowflake announced it had acquired Crunchy Data for about $250 million to bring a managed PostgreSQL service to the platform under the name Snowflake Postgres. (This enterprise Postgres adds an open source database engine to Snowflake’s AI Data Cloud). The timing makes sense when you consider PostgreSQL recently overtook MySQL as the top database in Stack Overflow’s 2024 developer survey.

But wait, there's more. Snowflake also unveiled its partnership with the LA 28 Olympics, Paralympics, and Team USA. As the official data collaboration provider, Snowflake will handle everything from athletes’ training data to fan engagement analytics.

Check out this summary of key highlights from Day 1 of the Snowflake Summit:

Day 2 — Tuesday, June 3, 2025: Big Product Announcements

Day 2 (June 3) turned out to be the most product-intensive day of Snowflake Summit 2025. Benoit Dageville and Christian Kleinerman took the stage for the Platform Keynote, introducing innovations designed to make enterprise data operations "easy, connected, and trusted".

⚡ Compute Performance Gets Real Improvements

The compute improvements received initial attention. Snowflake rolled out significant enhancements to its compute capabilities. Snowflake introduced Gen2 Standard Warehouses (Generally Available), which delivers about 2.1× faster query performance for core analytics. 

Alongside Gen2, Adaptive Compute entered private preview - an automated system that handles cluster sizing, auto-suspends and resumes clusters, and pools resources across workloads without manual intervention.

🤖 Conversational AI for Your Data Warehouse

On the AI front, Snowflake rolled out Snowflake Intelligence (public preview) – a conversational AI layer that sits on top of your data. Business users can ask natural-language questions about tables, documents, and spreadsheets, getting instant answers and visualizations. It is accessible via ai.snowflake.com.

Snowflake Intelligence is basically like ChatGPT, but for your enterprise data warehouse.

They also unveiled a Data Science Agent (private preview) to automate routine ML work (data prep, model tuning) for data scientists. These agentic AI features promise to "break down barriers" so anyone in an organization can gain insights from all their data.

🔗 Data Integration Updates

Data integration got a boost with Snowflake Openflow (general availability on AWS), a multi-modal data ingestion service built on Apache NiFi. Openflow offers hundreds of ready-to-use connectors for moving data from anywhere into Snowflake.

Snowpipe Streaming also entered private preview, offering high-throughput, low-latency streaming ingestion into Snowflake.

🔒 Data Security and Governance Updates

Security and governance also saw significant updates. Snowflake announced many governance enhancements. Snowflake Horizon Catalog got smarter. Horizon Catalog can now federate across Apache Iceberg tables (via "Catalog-linked databases") and will support external-data discovery (preview soon). A new Copilot for Horizon Catalog (private preview) was also announced, letting users use natural language to search and govern data directly in Snowsight.

On security, Snowflake extended its Trust Center (GA soon) and added new MFA methods and passkey support, as well as automated leaked-password protection (GA) and bad-IP blocking (GA soon) to strengthen account security. Snowflake also announced immutable Snapshots (point-in-time backups, in preview) for stronger data recovery, and its real-time Observability Tools are now generally available to give teams unified diagnostics and alerts. 

📊 Analytics Tools for Modern Workloads

On the AI-powered analytics side, Snowflake introduced Cortex AISQL, letting analysts embed AI operators directly in SQL for multimodal data queries.

They also announced SnowConvert AI, which is a free, AI-driven tool to accelerate migration from legacy platforms using AI-powered automation, reducing manual re-coding and lowering the risk associated with large-scale data migrations.

Together, these aim to make analytics and data migrations faster and cheaper.

🛍️ Marketplace and Agentic Apps

Finally, on the Marketplace and agentic apps side, Snowflake expanded its Marketplace with new "agentic" offerings. Cortex Knowledge Extensions (coming soon) let AI applications pull in curated news and research content to enrich AI insights.

Snowflake also previewed the sharing of Semantic Models (AI-ready data definitions) from partners, and launched Agentic Native Apps in Marketplace – third-party AI apps that run inside your Snowflake environment. These products make it easier to build AI-powered apps and agents on top of Snowflake’s data cloud.

Snowflake also made other significant key announcements, including Defining Semantic Views (GA) and Querying Semantic Views (preview) for mapping business metrics to raw tables, Snowpipe Streaming’s new high-performance pipeline (preview), and several Snowflake Native App Framework updates.

Check out this summary of key highlights from Day 3 of the Snowflake Summit:

All told, Day 2 was packed with technical detail. In brief: Snowflake's data platform will automatically tune compute. It will bring any data into Snowflake through Openflow and Catalog-Linked DBs. AI will be integrated everywhere, from analysts’ SQL to data scientists’ workflows. This all happens while preserving enterprise-grade governance.

Day 3 — Wednesday, June 4, 2025: Customers, Partners, and Breakouts

Day 3 (June 4) shifted focus from product announcements to practical applications. Customer breakout sessions and partner showcases dominated the agenda, showing how these new capabilities work in practice

The partner ecosystem expanded with Snowflake Ventures investing in Sema4.ai, a startup building AI agents for enterprises. The Sema4 Team Edition agent will be available as a Snowflake Native App, allowing customers to deploy AI agents directly within their Snowflake environment.

A highlight of Day 3 was the Snowflake Partner Awards. Snowflake honored 30+ partners for outstanding solutions and services.

Here are all the notable winners of the Snowflake Partner Awards:

🔮 Data Cloud Services Partners

• Global Implementation Partner of the Year: Cognizant
• Global Innovation Partner of the Year: Kipi.ai
• Global AI Partner of the Year: Slalom

🔮 Industry Data Cloud Services Partners

• Financial Services: EY
• Retail & CPG: Tredence
• Healthcare & Life Sciences: Hakkoda (an IBM Company)
• Public Sector: Deloitte

🔮 Data Cloud Product Partners

• AI Product Partner of the Year: Dataiku
• Data Integration Product Partner of the Year: Qlik
• Business Intelligence Product Partner of the Year: Sigma
• Data Security Product Partner of the Year: Securonix

🔮 Industry Data Cloud Product Partners

• Financial Services: S&P Global
• Retail & Consumer Goods: Blue Yonder
• Healthcare & Life Sciences: IQVIA

Check out the full list of winners and all event partners below 👇

These awards recognized a range of regional leaders and technology partners across AI/ML, data integration and analytics. (Many Snowflake OEMs and partners also had booths on the expo floor during the Summit).

After all the session was over, several deep-dive demos were shown: walking through Snowflake Intelligence’s conversational UI, configuring a Cortex AISQL query over images, using the Snowflake AI Observability tools, deploying Openflow pipelines, etc. (Snowflake also previewed integration with external tools – e.g. a joint solution with Oracle for CDC – demonstrating cross-platform data movement).

And in addition to Sema4, Snowflake showcased new agentic offerings on Marketplace: live demos of agentic apps by providers like Genesis Computing, S&P Global, and others (enabling data teams to install AI-powered assistants from the Marketplace).

Beyond awards, hundreds of sessions covered AI/ML, Snowpark (Python/R), data engineering, and Snowflake app development. Hands-on labs let attendees try Cortex AI SQL queries, build Snowflake-native applications, and define new Semantic Views (enterprise metrics models for AI) in Snowsight.

Day 4 — Thursday, June 5, 2025: Developer Day, Demos & Wrap-Up

Day 4 (June 5) was Dev Day – a free event for engineers, data scientists and developers. It featured hands-on labs, expert sessions, and community talks on AI, machine learning, and data engineering.

🛠️ Developer Tooling Updates

Snowflake rolled out new developer-centric features. Native "Workspaces" were demonstrated (environments for SQL and dbt development with AI code assistance). The GA Terraform provider was highlighted, making it easier to manage Snowflake resources as code. Snowflake also called out Snowpark Container Services for deploying custom ML models and agent code (especially with AWS/GCP integration in preview).

👥 Community & Training

Dev Day had numerous labs on Snowpark Container Services, Streamlit Apps on Snowflake, and Snowflake ML. Developers could earn certifications, meet Snowflake engineers, and join a post-event coding party. (Snowflake promoted its Data Superheroes community and user groups, and encouraged contributions to Snowflake’s open source projects).

🤝 Customer/Partner Sessions

At the very final customer/Partner Sessions, there were several case-study talks: for example, Salesforce data teams showed how they use Snowflake APIs, and Splunk/Snowflake partnership engineers talked about modernizing event pipelines. (The vibe emphasized that "builders" can leverage everything announced this week in their own apps and workflows).

🏆 Snowflake Startup Challenge Finale

One of the big events was the Snowflake Startup Challenge Finale. Three startups (DeepTempo, Lumilinks, and Winning Variant) pitched their products on a live stage. Snowflake’s CEO Sridhar Ramaswamy and judges from Conviction VC and the NYSE were in the audience. Lumilinks won the $1M grand prize (investments and mentoring). Lumilinks had impressed judges with its "FleetSense AI" solution.

Finally, the summit wrapped up with an eye to the future of Snowflake’s community. Snowflake shared its plans and urged attendees to connect on its developer channels. All in all, the four-day event gave a complete picture of what's happening with Snowflake. Platform keynotes highlighted new features like improved computing power, AI, and governance. Builder sessions showed how to put those features to use. Meanwhile, Dev Day showcased Snowflake's expanding network of labs, partners, and startups.

Conclusion

And that's a wrap! This year’s Snowflake Summit was packed with features and drew a large crowd. The main theme was AI and Snowflake positioning itself as the core infrastructure layer for enterprise AI. There were many announcements. The biggest was Snowflake acquiring Crunchy Data for $250 million. Another highlight was the partnership with LA 28 Olympics, Paralympics, and Team USA.

Snowflake is moving beyond traditional data warehousing. The new features cover conversational AI, automated data science, and agentic applications. The focus is on making advanced technology easy to use.

The summit showed a clear shift from vision to real products. Snowflake is building a strong technical foundation for enterprise AI.

These new features and capabilities are impressive; that's not in question. What is, is whether large enterprises can implement them without the complexity that typically comes with rapid change.

Additional Resources

• Snowflake Summit 2025 Platform Keynote
• Summit 2025 | Day 1 Highlights
• Snowflake Summit 2025 Opening Keynote With Sridhar Ramaswamy, Sam Altman, And Sarah Guo
• Summit 2025 | Day 2 Highlights
• Summit 2025 | Day 3 Highlights
• Snowflake acquires Crunchy Data
• Canva Accelerates Global Growth and Product Velocity with Snowflake AI Data Cloud
• Snowflake partnership with LA 28 Olympics, Paralympics, and Team USA
• Snowflake Unveils Next Wave of Compute Innovations
• An Even Easier-to-Use and More Trusted Platform from Snowflake
• Snowflake Introduces Cortex AISQL and SnowConvert AI
• Snowflake Openflow
• Snowflake Horizon Catalog
• Snowflake Immutable Snapshots
• SnowConvert AI
• Snowflake Workspaces
• Snowpark Container Services AWS/GCP integration
• Snowflake Summit 2025 Highlights: Building the Future of AI and Apps