This is a major product release for Cognite Atlas AI, which enables companies to seamlessly build and deploy AI Agents at scale to automate their complex operational workflows.
The latest release of Cognite Data Fusion® is a significant breakthrough for industrial operations and delivers the foundation for AI-powered autonomous operations by deploying AI agents at scale. The latest advancements of Cognite Atlas AI™ simplify AI agent building with an easy-to-use interface and ensure high-quality agents by integrating them with customers’ industrial knowledge graph, minimizing the risk of hallucinations and data inaccuracies.
Our vision for Cognite Data Fusion® is to enhance industry efficiency through autonomous operations, with AI playing a crucial role. We achieve this by providing simple access to complex industrial data, enabling scalable use cases with both out-of-the-box and tailored solutions, and leveraging AI agents to automate workflows. With this release, we are revolutionizing how industrial teams build and use AI agents to accelerate development cycles, reduce costs, and enhance usability, unlocking more value from AI initiatives.
Knut Vidvei, Head of Product Management at Cognite
Fully integrated setup for AI agents
This release brings all Cognite’s customers the power of generative AI to their industrial operations by providing a comprehensive setup that lets customers tap directly into the power of foundational generative AI models, benchmarks these models for superior performance, and minimizes the risk of hallucinations with Context Augmented Generation (CAG).
AI agents built with Cognite Atlas AI™ can be used directly in Cognite Data Fusion® via industrial tools like Industrial Canvas and Charts, customers' proprietary applications, and Cognite partner applications.
Rapidly innovate with an easy-to-use AI agent builder
Building AI agents for industrial applications can be complex and may often require deep AI and data science expertise. To simplify this, at the heart of Atlas AI is a low-code agent builder experience with ready-to-use language models, specialized AI agent tools for interacting with the industrial knowledge graph, and agent templates to jumpstart development. These tools reduce the need for specialized expertise, enabling faster development cycles and empowering more users to create and manage AI agents efficiently.
Utilize the best language model for each use case
The rapid release of new generative AI language models makes it challenging to select the best fit for each AI Agent. To streamline this process, Cognite Data Fusion® now offers direct access to generative language models from various providers, complete with benchmark results tailored to each industrial use case. With easy access to top-tier models, companies can significantly reduce the time spent selecting the optimal language model when developing an AI Agent and ensure trusted and reliable results.
Easily deploy and interact with AI agents
Cognite makes it easy for users to find, start, and interact with AI Agents directly within industrial applications like Industrial Canvas and Charts or through the AI Agent library. This provides an integrated experience for users with seamless data handling and analysis using natural language directly in Cognite Data Fusion®.
In addition, all users of Cognite Data Fusion® can now use document question answering, document summarization, and semantic search capabilities. These features are accessible for single and multiple documents in Search, Industrial Canvas, and Infield, with tooling available for Atlas AI agents. This enhancement reduces the time needed to retrieve insights and improves the quality of answers by combining relevant information from multiple documents.
Cognite's industrial applications get even better
With this release, Cognite is also enhancing its industrial applications, specifically Search and Industrial Canvas, to allow users to be even more productive in their day-to-day:
- A high-performance Industrial Canvas: Performance improvements and an improved user experience make root cause analysis (RCA) and troubleshooting processes 3x times more efficient. Improvements include reducing document loading times from 5 minutes down to just 20-30 seconds, displaying specific document pages directly, toggling time series data views, supporting clickable URLs, and customizable PDF exports.
- One place to search for all data: Users can now explore asset-centric and knowledge graph data in a single place with document categorization, search and filter in a tree view, file download, and viewing indirectly linked data.
- Collaborate directly in 3D models: Users can now mark, comment on, and share specific locations in 3D scenes, capturing details like coordinates and camera views. This boosts productivity by reducing search time and streamlining team workflows.
- Even better 3D navigation: Ghosting for 3D content, contextualized 360 images, navigation between previously viewed 360s, and toggleable markers improve the clarity and discoverability of 3D data.
A superior data foundation to enable AI agents
To enable AI Agents, Cognite keeps improving the foundational features of the industrial DataOps platform. This release improves the ability to ingest, contextualize, simulate, and govern industrial data to power the next wave of operational use cases.
- Detailed Diagram Parsing: This new feature detects symbols, tags, lines, and connections in vectorized diagrams to construct a knowledge graph to enable automated isolation planning and data validation.
- Better connectors and data retrieval: Up to 10x times faster data retrieval with greater flexibility and scalability, and new features that include Cognite Authentication Service support, REST API functions (GET and POST) for accessing Cognite Data Fusion® resources, and GraphQL for efficient queries.
- Advanced API-based simulation workflows: Users now have a vendor-agnostic service with out-of-the-box connectors and a simulator API for easy integration and flexible input/output handling and processing of simulation data.
- Better data workflows and enhanced SAP integration: Data is kept up to date via event triggers in data modeling, which include batching and starting workflows based on challenges in data modeling instances. Additionally, we made multiple updates to better support the connection to SAP, including SAP write-back.
Cognite is the leader in industrial AI
The December 2024 release of Cognite Data Fusion® with Cognite Atlas AI™ cements Cognite's position as the leader in industrial AI by delivering groundbreaking advancements that simplify AI agent creation and unlock the full potential of generative AI for industrial workflows.
With its intuitive low-code AI agent builder, seamless integration into advanced industrial applications, and a robust industrial data foundation, Cognite continues to set the benchmark for transformative AI-driven solutions.
For more information or questions, please get in touch here or dive deeper into the December 2024 Product Release with the below materials that include selected deep dives: