Snowflake has expanded its CoCo coding agent and launched Datastream, a managed streaming service for Apache Kafka. The moves deepen Snowflake's push into AI software development and real-time data processing.
CoCo, previously known as Cortex Code, is designed to help users automate workflows, build applications, and work with enterprise data through text prompts. The latest update adds a desktop application, mobile access, a Slack interface, and integrations with tools including Microsoft Excel, VS Code, and Claude Code.
The product now supports more autonomous task execution, allowing users to set recurring or event-driven workflows without repeated prompting. Snowflake has also added cloud-based agents that can run tasks in the background from Snowsight, along with a local sandbox designed to isolate agents from sensitive files and system resources.
Snowflake also introduced a Skill Catalog and pre-built Skills for data engineering and AI workflows. These are intended to let teams reuse workflows across functions rather than rebuild them for each project.
Broader reach
The update reflects a broader race among software companies to embed AI assistants into development tools used by both engineers and non-specialist staff. Snowflake is positioning CoCo as a way for analysts, business users, and developers to create data pipelines, automations, and AI applications within a governed environment tied to its data platform.
Christian Kleinerman, Executive Vice President of Product at Snowflake, described the company's view of the shift. "CoCo makes work for experienced developers dramatically faster and easier, and is opening the door for non-traditional builders, like analysts and data-savvy business users, to start creating pipelines, automations, and AI apps on their own," Kleinerman said.
He added: "With Snowflake Datastream bringing fresh, continuously flowing data directly into Snowflake, organizations can now build and operationalize real-time AI apps without the complexity of managing separate streaming infrastructure. When building with AI becomes as simple as describing the outcome you want, the number of people who can contribute to an organization's AI strategy increases by orders of magnitude and accelerates the shift to the agentic enterprise."
Snowflake said CoCo is being used by Fanatics, Thomson Reuters, and WHOOP. Those customers described the product as a way to cut the time spent on data engineering, troubleshooting, and workflow automation.
At Fanatics, the tool is being used to manage changing data demands and pipeline work. "At Fanatics, our data demands shift constantly, and Snowflake CoCo gives our team the speed to keep pace," said Maddy Want, Vice President of Data at Fanatics. "Engineers who used to spend days untangling pipeline issues and modeling data can now resolve those problems in hours, freeing them to build and ship new capabilities exponentially faster. Because it's integrated with our governed data, we're able to turn insights into action more quickly, powering everything from audience segmentation to real-time fan engagement. CoCo is helping us accelerate innovation across the business and deliver the personalized experiences our fans and partners expect."
Thomson Reuters is applying the tool on top of a Snowflake data estate spanning more than 37,500 governed tables and 350 data sources. "Thomson Reuters has built its data foundation on Snowflake to create a single source of truth across 37,500+ governed tables and 350 data sources, and now Snowflake CoCo is accelerating how we build on top of that," said Caitlin Halferty, Head of Data & Analytics at Thomson Reuters. "Our teams are modernizing legacy systems, scaling AI pipelines, and delivering insights in days instead of weeks, all within a governed environment. When you can go from idea to production that fast on top of trusted data, it fundamentally changes what's possible."
WHOOP plans to roll out CoCo beyond its data team. "Snowflake CoCo is becoming a core part of how we operate as a company," said Matt Luizzi, Vice President of Analytics at WHOOP. "We're rolling it out across the entire organization, not just our data team. Because it understands our data, our environment, and our governance, teams can build and automate workflows on top of trusted data without needing specialized expertise. What used to take multiple systems and manual coordination can now be done in a single, governed environment. CoCo is helping us move faster, reduce complexity, and embed AI into how work gets done every day."
Streaming service
Datastream addresses a separate but related issue for companies building AI systems: how to move live data into analytics and application environments without running separate streaming infrastructure. The Kafka-compatible service lets organizations stream data directly into Snowflake tables while applying the same governance, access controls, masking, and lineage rules used elsewhere on the platform.
The approach is aimed at reducing the need for brokers, connectors, and duplicated permissions across different systems. For businesses building AI applications that depend on current operational data, the appeal is less about raw model development and more about simplifying the movement of live data into governed stores.
Snowflake is also tying Datastream directly to CoCo. Users will be able to create and operate real-time pipelines through prompts, combining live data ingestion with AI-assisted development in the same environment.
The launch broadens Snowflake's effort to make its platform the place where customers store data, govern it, and build AI software on top of it. It also shows how vendors are trying to bring software development, workflow automation, and streaming data operations closer together rather than leaving them in separate tools and teams.
According to Snowflake, Datastream is built to connect existing streaming applications and data flows without requiring organizations to redesign their systems.