Autonomous master data governance agents

Autonomous agents can work through large sets of records in the background with Long Running Tasks, so data stewards can focus on the records that need review. Custom agents can also run as autonomous jobs.

The following capabilities are available:

• Start from a conversation or a schedule: Start a job from a prompt or an attached CSV, JSON, or XLSX file, or schedule it in Task Management to run now, once, or on a recurring schedule.

• Review results for each record: Each record gets a status, a confidence score, a reason, and a detailed trace. Records the agent is uncertain about are marked Needs review for data stewards. Export the results as CSV or JSONL from the user interface or through an API.

• Run custom agents as autonomous jobs: Authors turn on Long Running Tasks as a capability of a custom agent, and a feasibility check confirms that the agent is suited to it.

• Isolated, scalable runtime: Each customer's jobs run on a dedicated, isolated runtime, so they don't compete with other tenants. Scaling and performance improvements let agents process larger record sets faster and more efficiently.

Reltio enables Long Running Tasks once for each tenant. No data model changes are needed.

Unstructured search

Unstructured search is a new capability in Unstructured augmentation (earlier known as AgentFlow Unstructured) that turns documents from supported sources into AI searchable knowledge bases. You can configure and test knowledge retrieval with a sample document before indexing the full set of documents. Enable this feature on custom Reltio agents for them to search and provide answers grounded in relevant organizational content.

Unstructured search reduces the need to manually search policies, contracts, product documentation, support runbooks, and other enterprise content. Custom agents can use one or more knowledge bases alongside governed Reltio data to support data unification decisions, customer service, compliance reviews, and operational questions.

New velocity packs for faster implementation

Velocity packs are pre-built, ready-to-deploy configurations that give you a working starting point for your Reltio tenant. Instead of designing a data model, matching rules, and related settings from scratch, you start from proven, industry-aligned content and adapt it to your business. This reduces implementation effort, shortens time to value, and brings consistency to how new tenants are set up.

Reltio now delivers a new set of velocity packs that cover a broad range of domains and use cases. Each pack gives you a pre-configured data model to start from, so you can reduce design and configuration work, adopt a consistent approach across tenants, and adapt the content to your business by keeping, extending, or hiding what you need.

Docs for You

Reltio Docs for You is an AI-powered application that turns official Reltio documentation into a personalized guide. You provide your role, your goal, your learning style, and your experience with Reltio, and the application selects the most relevant documentation topics and generates a guide that includes prerequisites, sequenced steps, key term definitions, and links to the source topics. This helps data stewards, administrators, developers, and other roles reach their goal without manually piecing together topics from search results.

Each guide is tailored to the learning style you select. You can regenerate it, share a link to it with colleagues who have Reltio credentials, download it as Markdown, or print it as a PDF. Docs for You is not a chatbot and doesn't connect to your tenant. Any Reltio customer with a tenant ID and valid user credentials can use this application.

Find it in the top-left corner of the Documentation Portal or go straight to the Docs for you application.

Materialized hierarchy management including attribute inheritance

Materialized hierarchy helps organizations create, govern, and maintain complex business structures for customers, products, territories, organizations, and financial data.

Organizations can maintain current, future, and historical hierarchy versions while preparing changes without disrupting the hierarchy used by current business operations. Structural validation helps prevent invalid hierarchy changes from reaching downstream consumers. Attribute inheritance applies common business values through a hierarchy, while allowing overriding a local value for individual nodes or branches.

The following capabilities are available:

• Named hierarchy versions with temporal validity: Maintain multiple hierarchy versions with defined validity periods to support current, future, and historical business structures.

• Lifecycle management and activation: Manage hierarchy versions over time and activate the appropriate structure based on business timelines.

• Structural validation: Apply hierarchy rules and validations to help prevent structurally invalid changes from propagating to downstream systems.

• Attribute inheritance and overrides: Apply common attribute values through hierarchy levels while allowing to override a local value for individual nodes or branches.

• Hierarchy replacement during import: Replace the complete content of an existing hierarchy or version when imported names match. The hierarchy instance ID is retained, connection IDs may change, and unmatched names create a new hierarchy or version.

• Advanced hierarchy maintenance: Manage complex, multi-level structures with drag-and-drop sibling ordering, identify hierarchy nodes using crosswalks instead of entity ids, and receive notifications about hierarchy issues.

Flexible entity resolution network (FERN)

Flexible Entity Resolution Networks (FERN)™ provides more control over the profiles it evaluates, how it works with customer-defined match rules, and how match confidence is interpreted. FERN also supports inference through Google Cloud Vertex AI, Amazon SageMaker, and Azure Machine Learning, and processes multivalued attributes more efficiently.

SAP Integration Suite Adapter for Reltio

The SAP Integration Suite Adapter for Reltio provides native connectivity between SAP Integration Suite and Reltio, replacing custom HTTP/REST implementations with consistent, managed connectivity and reducing implementation time.

Zero-copy interaction integration from Google BigQuery

Zero-copy Interaction Integration from Google BigQuery allows Reltio Intelligent 360™ to access and use interaction data stored in BigQuery without replicating data into Reltio, connecting securely through Google Cloud credentials and querying data in place while linking it to Reltio profiles through crosswalks.

This capability brings interaction data stored in Google BigQuery into Reltio Intelligent 360™ profile views, where it appears alongside mastered profile data in the Hub. Querying the data in place avoids replication, helps maintain data governance, and reduces the work of managing another copy. It also lets the interaction timeline reflect near-real-time activity available in BigQuery, enabling more accurate customer insights.

Enhancements to 360-degree entity profiles

Integrated development experience

Integrated development experience is an AI-powered development experience for VS Code and Cursor that helps teams build, manage, and deploy Reltio business configurations faster and with greater confidence. By connecting to Reltio tenants and Git repositories, it brings AI-assisted authoring, intelligent navigation, real-time validation, and ontology visualization into familiar development workflows. Within this experience, teams can collaborate on updates, review differences before deployment, and trace changes through configuration history.

Integrated development experience now supports workflows and Life Cycle Actions (LCAs), helping teams develop and deliver custom business logic efficiently. Teams can create workflows from a starter template or duplicate existing ones, visualize their structure, and edit BPMNs with live validation. From there, they can compare changes before deploying to a tenant or committing to a Reltio-provisioned Bitbucket repository, with deployment history available to track subsequent updates. For teams working with LCAs, the experience extends from exploring existing LCAs to generating LCA code through AI prompts and deploying it to their own AWS, GCP, or Azure environment. To help bring these functions into use, integrated development experience guides and updates the tenant's business configuration to link the deployed function.

Data sharing - essentials

Reltio data sharing - essentials makes trusted Reltio data available in supported cloud data platforms, such as Databricks and Microsoft Fabric through the target platform's native sharing mechanism, for analytics, AI and data quality use cases. Each tenant supports a single active All Values data share, giving downstream teams access to operational and non-operational values along with crosswalk context.

With this added context, teams can better understand where values came from, what other values existed, and which source systems agree or disagree with the trusted value. Data synchronization is triggered every four hours, with updates becoming available in the data share after queued events are processed, helping customers avoid data movement while enabling governed downstream consumption of core Reltio data objects, including entities, relationships, interactions, matches, merges, and links.

Reltio AgentFlow is now Autonomous agents

Autonomous agents is the new name for AgentFlow, and each capability now has a name that describes its function.

Self-service metadata security management in Reltio Console

Metadata security controls which roles can access entity types, relationship types, interaction types, and individual attributes in your tenant. Managing these permissions required hand-editing JSON permission files and calling the Permissions Configuration APIs, so routine changes often depend on technical users or Reltio Support. This slows down configuration, increases operational effort, and makes it hard to confirm who can access a resource or record.

Reltio now provides a Metadata Security app in Console, so authorized administrators can manage role-based permission rules through a guided user interface.

Zero-copy segmentation from Databricks

Zero-copy segmentation with Databricks extends segmentation in Reltio Intelligent 360 to interaction and transaction data stored in your Databricks workspace. You can create and manage segment rules using profile and interaction attributes without moving interaction and transaction data into Reltio, so the data remains governed in Databricks.

Enhancements to Custom agents

Custom agents (earlier known as Agent Builder) now support Long Running Tasks which let an agent apply the same repeatable work across a set of records. When you enable Long Running Tasks, a suitability check confirms that your instructions and setup describe a clear record-processing task.

OAuth authentication with Azure Entra ID for Confluent Kafka external queues

Confluent Kafka external queues connect Reltio to Confluent Cloud so that master data change events can be streamed to Kafka topics for downstream consumption. Confluent Cloud environments secured with Azure Entra ID rely on OAuth 2.0 authentication to establish these connections.

Reltio now supports SASL/OAUTHBEARER authentication for Confluent Cloud using the OAuth 2.0 Client Credentials flow with Azure Entra ID. Reltio obtains the required access token using the configured OAuth credentials and refreshes it before it expires, helping maintain continuous connectivity to the Kafka cluster.

This support enables customers using Azure Entra ID to stream Reltio events to Confluent Cloud while continuing to use their existing identity and access management setup.