AI Innovation with Seamless Collaboration

Let’s face it: If you’re not thinking about how to incorporate AI into your products and your internal processes, your competitors — and potential customers — are. The ability to quickly harness AI’s transformative potential may mean the difference between market success or stagnation. And the key to shifting from AI experimentation to delivering real business value lies in how you leverage your data — how well you can share it, find it, access it, build with it, monetize it and use it. 

With that in mind, Snowflake is aiming to redefine how organizations leverage data and AI with a slew of enhancements that make it easier to connect to and collaborate with external resources and internal teams, extend your Snowflake platform with previously out-of-reach AI and analytics capabilities, and build and deploy AI-driven apps while protecting private data.

Removing commercial barriers to AI

Snowflake is launching new features that remove commercial and integration barriers to using third-party data in AI applications and agentic systems. These innovations enable enterprises to easily connect and contextualize their own first-party data with external knowledge sources — including structured and unstructured data — in a trusted manner.

Agentic Snowflake Native Apps on Snowflake Marketplace

Snowflake now supports Agentic Snowflake Native Apps on Snowflake Marketplace, a new way for providers to develop and offer Snowflake Native Apps that reference Cortex Agent APIs. These interoperable agentic products can give data engineers, data scientists and others a fast track to deploying and creating value from agentic AI. Providers may build third-party Agentic Native Apps with anything from pre-packaged data and pre-trained ML models to fine-tuned LLMs; the apps can be used within Snowflake Intelligence or as a building block in other agentic apps built on Snowflake Cortex AI. 

Because Agentic Native Apps can be brought to the customer’s proprietary data and run in the customer’s Snowflake instance, there’s no need to copy or move data to reap the benefits of agentic AI. This allows your teams to quickly incorporate third-party agentic experiences into their applications, enrich agentic experiences that involve Snowflake Intelligence, and accelerate development and monetization of AI. Where providers want to develop Agentic Snowflake Native Apps to orchestrate or execute tasks, they may leverage or work in concert with Snowflake Cortex Agents. 

Cortex Knowledge Extensions and Sharing of Semantic Models on Snowflake Marketplace

Snowflake Cortex Knowledge Extensions (generally available soon) allow your organization to easily talk to third-party unstructured data from media and content publishers, such as The Associated Press, Washington Post, USA TODAY Network, Packt, Stack Overflow and CB Insights into AI assistants and agentic systems directly from Snowflake Marketplace, all while respecting the intellectual property of publishers and commercial terms and supporting proper attribution. Explore all the Cortex Knowledge Extensions now available on Snowflake Marketplace and chat with the official Snowflake docs now through the Cortex Knowledge Extension — check out Snowflake Documentation.

Snowflake customers can now also easily integrate AI-ready internal data sets or third-party structured data in their AI assistants and agentic systems with sharing of Semantic Models (private preview), which removes the need to spend time and effort on semantic model creation for shared data. Whether the data is shared by internal data owners, SaaS providers or commercial data providers, your teams can now use natural language to “talk to data” — including data from providers such as CARTO, CB Insights, Cotality powered by Bobsled, Deutsche Börse, IPinfo and truestar — because the providers have already created the semantic model for their data sets.

With these AI-enabling technologies, you don’t need to define additional semantics to incorporate that data into Snowflake Cortex-enabled applications such as Snowflake Intelligence. This makes it easier to enrich AI assistants and agentic systems with external structured or unstructured data that would typically be safeguarded behind paywalls to protect the IP of the data providers.

Making it easier to purchase and onboard data products from Snowflake Marketplace

The Marketplace Capacity Drawdown Program is expanding, giving Snowflake customers even more flexibility in how they use their Snowflake credits. Eligible customers can now use committed Capacity to purchase offerings from integrated AI Data Cloud Product Partners such as dbt, Immuta, Monte Carlo, Sigma, Matillion and more. By consolidating your data and AI-related spending under your Snowflake contract, you can avoid additional budget approvals and vendor vetting and speed up the time it takes to start working with the solutions you need. (Program restrictions apply; see documentation for details.) 

With the introduction of Offers on Snowflake Marketplace (public preview soon), Snowflake customers can also negotiate custom terms and pricing specific to their organization’s needs when purchasing partner products on-platform. Offers also allow more partners to list their solutions on Snowflake Marketplace, while improving ease of use, controls and speed when closing highly negotiated deals on Snowflake. 

Furthermore, Snowflake Native App providers get even more optionality in the Monetization models offered. The commitment + usage model (private preview soon) simplifies pricing negotiations due to direct alignment with the Snowflake pricing model. Snowpark Container Services' time-based surcharge (public preview soon) aligns price to value based on the underlying Snowpark Container Services resource time utilization.

Raising the bar on Snowflake Native App security and privacy-enhancing, interoperable collaboration

Security is top of mind for everyone, and we’re taking steps to make it easy to build it into every app built on Snowflake. Our latest enhancements to the Snowflake Native App Framework were designed to improve security and interoperability.

The first improvement focuses on application versioning. To streamline CI/CD and release management, providers can now leverage powerful new features: 

• Simultaneous multiversion app deployments enable seamless coexistence for staged rollouts and A/B testing.

• Custom metric emission directly from app code provides enhanced observability.

• Session debug mode offers granular control over app objects and execution, simplifying troubleshooting and development.

Providers can now deploy both generally available and pre-release versions of an app concurrently to a customer, helping to smooth transitions and feature testing. This helps streamline user acceptance testing and simplifies performance tracking and troubleshooting, providing greater observability and accelerating development of reliable applications.

Snowflake Native App permissioning is also getting an upgrade. Providers can now more easily configure access to their apps — both for themselves and for app users — by leveraging automated privilege grants, role-based access controls and user-context operations. These capabilities help teams reduce setup complexity and avoid extra provisioning, align with customer security models, and provide access to database objects with more granularity, all while maintaining strong administrative governance. 

Users and developers of Snowflake Native Apps are also getting better app observability with the metrics tab in Snowflake Trail. The metrics tab shows the apps’ CPU/GPU and memory usage along with key app events such as service fails, restarts, upgrades and even custom metrics logged from the app — all of which makes it easier to track down and troubleshoot performance issues.  

Customers who use Snowflake Native Apps from providers like Amdocs, Neo4J and NuSummit for next best action or fraud detection can now securely train the application’s model using their own data. By bringing the code and AI to the data, the data remains within the customer’s Snowflake instance, helping to reduce the risk of a security breach and enhancing prediction accuracy.

Snowflake is also making it easier for users to quickly identify Snowflake Native Apps that meet certain compliance requirements with the introduction of compliance badging. We’ve extended our cross-cloud reliability, as Snowflake Native App Framework integration with Snowpark Container Services is now in public preview on Google Cloud Platform. Our support for the major cloud platform providers helps customers reduce risk of regional outages when using third-party apps and data products: Snowflake failover capabilities mean the applications can continue working while minimizing risk of downtime. 

Democratizing data access without compromising data privacy

Making data available to more people — in a privacy-preserving way — is a valuable tool for driving progress and sparking innovation. Snowflake is committed to making collaboration within and between organizations seamless and simple by helping employees find, use and share the data resources they need without jumping through multiple hoops.

Snowflake Data Clean Rooms have become an essential tool for securely bringing data from multiple partners together to foster collaboration while safeguarding intellectual property and protecting assets. The advertising industry in particular has embraced these collaborative spaces, but ad measurement remains fragmented, particularly with the rise of streaming media. Advertisers and publishers struggle with tracking and attribution across multiple platforms and face limited visibility into campaign performance.

To address these challenges, Snowflake is driving continued innovation in data clean rooms by enabling a unified, privacy-aware collaboration environment tailored to advertising measurement needs. This orchestration capability integrates with Snowflake Data Clean Rooms to connect media publishers, advertisers and measurement partners in a neutral, flexible environment where they can securely share data and conduct custom measurements. 

By eliminating the need for rigid, one-size-fits-all solutions, this approach gives participants the flexibility to manage ad logs, define custom measurement logic and unlock consistent, real-time insights. This enables participants to conduct scalable, privacy-conscious ad measurement and optimize campaigns with precision.

Snowflake Horizon Catalog is also making it easier than ever for customers to derive value from sensitive data. The first step is discovery, which we’ve streamlined through enhancements to Horizon Universal Search that speed up and simplify the process of finding the data, apps and AI products you’re looking for. And Horizon Discovery will soon include third-party data discovery, helping users find and explore external assets without leaving the Snowflake platform.

Meanwhile, data governors can now enable secure discovery of sensitive data with the introduction of a request for access workflow within Snowflake Internal Marketplace. Data teams that lack the necessary privileges can simply look at the custom profile of a data product and request access, which not only encourages efficient cross-organization use of data assets but also facilitates data-driven collaboration.

And with the general availability of Egress Cost Optimizer, Snowflake makes global data sharing more efficient and cost effective. This feature helps data providers increase the availability of their data globally, reduce operational costs and gain more control over costs of cross-regional data sharing as they scale initiatives across organizations and business ecosystems.

Next steps

Want to learn more about these Snowflake enhancements and how they can help you build and deploy apps faster, connect to external resources and extend your Snowflake capabilities? Start here:

• Check out these Snowflake Summit 2025 sessions: Discover everything you need to know about connecting your business with zero-copy data sharing.

• Watch our data mesh demo: Learn how to build and govern your data products in Snowflake.

• Join the Snowflake Native App Bootcamp: Learn how to build, market, monetize and distribute apps to customers across the AI Data Cloud in this comprehensive 120-minute bootcamp.

   

Forward Looking Statements

This article contains forward-looking statements, including about our future product offerings, and are not commitments to deliver any product offerings. Actual results and offerings may differ and are subject to known and unknown risk and uncertainties. See our latest 10-Q for more information.