At Snowflake Summit 2026, Snowflake presented its 'Agentic AI' strategy, which integrates coding, business, and data governance, and emphasized the direction of transitioning enterprise AI from the experimentation stage to the execution stage, centering on Coco, Cowork, and the Horizon Catalog.

Snowflake held the 'Snowflake Summit 2026' in San Francisco, USA, from June 1 to 4 local time.

This event demonstrated that the paradigm of the data and artificial intelligence (AI) market is shifting in earnest to 'Agentic AI'.

At this event, Snowflake further advanced its 'AI Data Cloud' strategy by unveiling innovations across the board, including coding agents, personal work agents, data governance, and interoperability.

■ Coco and DataStream: Simplifying AI Development to an 'Explanatory Level'


One of the most notable announcements is the enhancement of the AI coding agent 'Snowflake CoCo'.

Coco supports everything from building data pipelines and automating workflows to developing applications using only natural language-based prompts, and is designed to enable not only developers but also non-developers to participate in building AI.

In particular, as integration with various environments such as desktop, mobile, Slack, VS Code, and Excel expands, it provides an environment where AI development is possible anytime, anywhere.

With the integration of 'Snowflake DataStream' for real-time data processing, a foundation has also been laid to simultaneously ensure the timeliness and accuracy of AI applications.

DataStreams significantly reduces operational complexity by enabling the direct collection and processing of Kafka-based data without the need for separate streaming infrastructure.

■ Cowork: The Evolution of Personal AI Agents for Knowledge Workers


'Snowflake CoWork,' a personal AI agent for knowledge workers, is also a key pillar.

Cowork is characterized by integrating data, tools, and work context to connect insights beyond deriving them to actual execution.

The deep research feature isIt comprehensively analyzes structured and unstructured data within the business to explain the 'causes' of insights, and supports business automation through user skills and multi-agent orchestration.

In addition, the Cortex training feature enables companies to train and optimize AI models using their own data, reducing reliance on costly external APIs and making it possible to build purpose-specific models.

This serves as a crucial factor in scaling AI from the experimental phase to actual operations.

■ Horizon Catalog: Securing Trust-Centric AI Governance


The focus was also placed on resolving the issue of 'trust,' a key prerequisite for the spread of AI.

Snowflake integrated data governance, security, and business context centered around the 'Horizon Catalog'.

The new 'Horizon Context' connects data across the organization using the same semantic system, enabling AI to make decisions based on consistent standards.

In addition, it enables data protection and control even in large-scale AI environments through agent identity, AI security posture management, and zero-trust-based security features.

The strategy is to structurally manage the risks that may arise in an environment where AI operates autonomously.

■ Open Interoperability: AI execution environment without data movement


In terms of data utilization, 'open interoperability' was presented as a key keyword.

Snowflake has enhanced Apache Iceberg-based technology to enable data utilization across various platforms and systems without data movement or replication.

This enables companies to manage data integrally under a single governance framework while flexibly operating multi-cloud and multi-engine environments.

In particular, data sharing has also evolved.

The shared data can be immediately utilized in the form of AI agents and is converted into 'agentic data' that derives insights through natural language queries.

This is an approach that transforms data from a mere asset into actionable intelligence.

■ Competition for 'Data + AI + Agent' Integrated Platforms Intensifies

Snowflake's message revealed at this summit is clear.

The goal is to connect data, AI, and applications on a single integrated platform and to expand the use of autonomous AI agents based on this.

As the 'Agentic Enterprise,' where AI assists or performs the entire process from coding and analysis to decision-making and execution, becomes a reality, the core of corporate competitiveness is shifting from the data itself to how quickly and reliably it is utilized.

Through this innovation, Snowflake is strengthening its role as a platform at the center.