What's New​

File Intake Agent 1.1 introduces the following updates:

Added​

• Added information about supported file types in the file upload UI.
• Added support for lowercase letters in entity names.

Changed​

• Renamed the query parameter memoryName to identifier.
• Improved path and file name concatenation to work independently of the operating system.

Fixed​

• Fixed an issue where error messages returned from DCH were not displayed.

What's New​

Version 3.1 of the DCH stack includes the following new features and improvements.

Agent Prompt Suggestion​

Agents now include a built-in prompt generation system that automatically constructs a structured, operational brief for the language model.

This prompt defines the agent’s objective, grounds it in the connected graph data (memories and target entities), and specifies a clear method and expected output format. Rather than a generic instruction, it acts as a context-aware execution blueprint.

Prompts are dynamically generated based on the agent’s connected memories, ensuring that every analysis is aligned with the actual data model, relationships, and available context.

What's New​

File Intake Agent enables file ingestion through both UI and REST API. The processing workflow:

• Files are uploaded and immediately forwarded to the File Extraction Worker for chunking
• Chunked content is stored in M4AI for semantic search and analysis
• File metadata (name, size, type, etc.) is persisted to the knowledge graph in DCH
• This integration contextualizes files within your data landscape, making both metadata and content fully searchable

What's New​

Version 3.0 of the DCH stack includes the following new features and changes.

Projections​

Projections introduce a unified approach to managing Views and Functions in DCH. Key capabilities include:

• Consolidated management: Views and Functions are now managed together under Projections
• Full versioning support: Create multiple versions and control their lifecycle through draft, published, and archived states—similar to M4AI agent systems
• Built-in testing: Validate Cypher queries directly in DCH before publishing, ensuring your projections work as expected

Additionally, the Views API has been relocated to the Linked Data API (formerly EBS) to better align with its focus on graph data operations.

What's New​

Intake Agents reach their first stable release, marking them as production-ready for data ingestion. They replace Data Pumps as the recommended approach for importing data into DCH.

Key advantages over Data Pumps:

• Language-agnostic development: Build agents in any programming language using a standardized REST API contract
• Decoupled architecture: Operate independently of DCH, simplifying updates and maintenance
• Enhanced performance: Process data in batches instead of loading everything at once, avoiding memory constraints
• Better observability: Batch processing provides immediate feedback and complete traceability throughout the import process

What's New​

Version 2.6 of the DCH stack includes the following new features and changes.

Exploration Session in GBS and Removal of Explorer​

The new Exploration Session feature has been introduced in the GBS UI, providing a user-friendly and intuitive way to navigate graph data. Users can now:

• Select data using labels.
• Search for specific data points.
• Explore data dynamically by adding related nodes and resolving paths between nodes.

With the addition of this enhanced feature in GBS, the Explorer and its backend are now obsolete and have been removed from the stack.

What's New​

Version 2.5 includes the following new features and changes.

Data Provider Workers (Beta)​

Data Pumps were previously the primary method for data import into DCH. This version introduces Data Provider Workers as the new system for importing data. A Data Provider Worker is a standalone application implementing a pre-defined contract and exposing a REST API for communication with one or multiple DCH instances. It can be developed in any programming language, offering enhanced flexibility. While Data Pumps are tightly coupled to DCH, the decoupled nature of Data Provider Workers allows for a smoother transition when changes are introduced.

Additional benefits include increased performance and traceability of imported data. Previously, Data Pumps loaded all available data at once, which could lead to memory constraints. Additionally, the single large job structure didn't provide visibility into processing progress. Data Provider Workers handle data in multiple batches, enabling quick feedback and comprehensive traceability.

What's New​

Version 2.4 introduces several key features and changes to improve stability, logging, and UI while preparing for future advancements.

Memory 4 Your AI (M4AI)​

Memory 4 Your AI is an innovative platform that leverages DCH to enhance pre-trained Large Language Models (LLMs) with graph-based contextual knowledge. It facilitates the development and management of specialized engineering agents, designed to provide domain-specific insights and automate complex, industry-focused tasks. This low-code tool enables the creation of virtual AI assistants with a deep understanding of context, terminology, and workflows, tailored to solve domain-specific challenges.

Version 2.4 also introduces a new UI (GBS Beta) which is set to replace Explorer and GBS UI in the future. In addition to supporting all M4AI features, the new UI includes a Model section that showcases target entities, relationships, and data sources offering a preview of the modern design and user experience DCH envisions for the future.

What's New​

Version 2.3 includes the following new features and changes.

New Explorer Backend​

A lot of effort was put into separating concerns of GBS (UI) and Explorer (formerly Frontend) to increase maintainability and stability. All Explorer-specific features were reevaluated and either removed or moved to a new dedicated backend. The following list contains all features that were removed.

• Data Lab provided basic data analysis tools that could be applied to nodes once Property Aggregation types were configured accordingly. It was also possible to define and attach an Interop Template to a Data Lab.
• Entity Configuration was used to define if an entity was used as scope, was a generic entity, could be used in sandboxes or if it should be versioned.
• Group node logic allowed to define a property name in system configuration that held a boolean value to define a node as group node.
  info

  To remove the now obsolete xpl_isGroupNode property from nodes, the graph has to be reprocessed.

• Integration enabled the definition of scripts that could be executed in Explorer.

What's New​

Version 2.2 includes following new features and changes.

Graph Security Layer​

New features to enable additional filtering and automatic Access Rule generation were added to Graph Security Layer.

Custom ExternalGraphSecurity workers allow to additionally filter graph related results after Access Rules are applied. Once registered in GBS, each graph related call to EBS will also call ExternalGraphSecurity workers which may then filter results based on customer specific systems or rules. For more information regarding the implementation of such a worker, refer to the developer documentation.

Since manually creating and maintaining Access Rules can be cumbersome, version 2.2 adds the possibility to automatically generate Access Rules based on external systems. A new Rule Transformation step can be added to Load Plans. This step allows to define a registered RuleTransformation worker that is called when the step is executed. Such a worker can be used to automatically create Access Rules based on customer specific systems. The data a worker needs to create rules (e.g., users, entities, properties, etc.) can be defined in the Rule Transformation step. For more information regarding the implementation of such a worker, refer to the developer documentation.