Cognite's latest product release simplifies the onboarding and orchestration of real-time industrial data with AI agents.
The latest release of Cognite Data Fusion® incorporates seamless interaction with AI agents in business workflows and data pipeline orchestration. Also included in this release is Data Workflows, a robust solution for increasing the scalability and observability of industrial data. Furthermore, the release automates the crucial process of contextualizing 360-degree images, streamlining data contextualization.
AI-powered data pipelines to increase observability and reduce data management efforts
We are pleased to announce that Data Workflows is now Generally Available for all customers. Data Workflows offers a graphical interface for building and managing tasks, embedding AI agents into these workflows to automate tasks, and tools for efficiently monitoring and inspecting workflow executions.
Onboarding high-quality, real-time industrial data to Cognite Data Fusion is essential for powering industrial applications and user experiences. Building data pipelines through APIs alone or setting up individual tasks can be prone to errors on larger scales and create friction for less technical users. The newly released Data Workflows provides a unified experience to move, transform, and manage data within an intuitive user interface. This allows users to link many different processing steps into one comprehensive job, delivering greater observability and reducing data management efforts. Throughout beta testing, over 2,500 workflows were created, with more than 150,000 daily runs.
Key capabilities in Data Workflows includes defining tasks for transformations, functions, simulations, other Cognite Data Fusion requests, and more. Users can create triggers for workflows to run based on schedules or events, such as new data in their data model and incorporate AI agents to automate processing tasks. The system also supports versioning to handle changes and iterations. Monitoring the history of runs helps users understand status and execution details, allowing for efficient debugging. Data Workflows simplifies the process of creating data pipelines and further democratizes efforts to manage industrial data.
Interact with multiple AI agents across your business workflows
Atlas AI agents are designed to accelerate daily work, from straightforward industrial data retrieval to complex use cases such as equipment troubleshooting and root cause analysis. This release introduces an improved experience. Users can now easily switch between available AI agents in a conversation without losing context. The system automatically selects the most recently used agent, further enhancing user convenience. Conversations persist across various user interfaces until a new session is initiated, ensuring continuity and context for users while executing tasks. for interacting with these agents. Additionally, native support for interacting with annotations significantly improves the experience when working with contextualized P&IDs.
Automate the contextualization of 360° images
3D models in Cognite Data Fusion can provide valuable information through 360-degree images of industrial facilities. Previously, users could inspect uploaded 360-degree images to locate equipment tags appearing as text and contextualize each one to the correct assets and equipment. This release reduces contextualization efforts by including automatic detection of text on equipment within 360-degree images. The system provides a list of suggested image areas likely containing tags for contextualization. This enhancement will significantly reduce manual efforts and accelerate the contextualization of equipment tags in large image collections.
To experience the power of our latest release, attend our upcoming product tour, Industrial AI in the Era of Contextualization and join us at Impact 2025 – the premier industrial data and AI conference. See firsthand how we're shaping the future of industry.