1. Oracle
  2. GoldenGate
  3. Data Fabric

Oracle GoldenGate Data Fabric

From business events to trusted actions.Turn committed operational changes into governed, traceable, AI-ready data products. Keep applications, analytics, and AI connected to your business as it happens—across the platforms you choose. GoldenGate Data Fabric is a unified product experience that turns GoldenGate's real-time foundation into governed data products that can be developed, operated, and consumed across distributed environments.

Take the product tour
One fabric across your operational, analytical, and AI estate
One fabric across your operational, analytical, and AI estate

What's new in GoldenGate Data Fabric?

  • From business events to trusted actions

    Get the architecture, the use cases, and a practical starting point for your first governed data product.

  • What is Oracle GG Data Fabric?

    See how one real-time fabric connects the systems that run your business to the decisions that move it forward.

  • Meet the new GoldenGate Data Fabric

    Discover a unified Workbench, governed Datasets, AI-assisted design, and open execution choices.

Why Oracle GoldenGate Data Fabric?

One real-time data fabric. Three outcomes. Connect how your business operates with how your enterprise learns. Build on the committed changes already flowing through your systems to make every next product and decision more useful.

Operate

Keep the business running.

Synchronize mission-critical systems through migrations, upgrades, and disruption. Extend proven GoldenGate replication across your distributed estate.

Activate

Make every change more valuable.

Capture committed change once and publish current, governed products for applications, lakehouses, analytics, and AI. Give each consumer data it can use.

Optimize

Keep your next choice open.

Separate data-product design from a single platform. Place supported workloads where performance, locality, sovereignty, and economics work best.

30+ years - Built on GoldenGate’s experience keeping mission-critical data synchronized, migrated, and recoverable across Oracle and supported non-Oracle systems.

The economics of choice

Standardize the fabric, not the platform.

Modern enterprises span operational databases, applications, clouds, and data platforms. GoldenGate Data Fabric connects them with current, governed data products while preserving where you store data and how you process it. Reuse the design. Keep processing close to the data. Make your next platform decision on your terms.

Learn more about Oracle GoldenGate source and target certifications
Diagram of Oracle GoldenGate Data Fabric and supported data platforms

Features and capabilities

Build the data product. Bring the trust with it.

Oracle GoldenGate Data Fabric is a managed service providing a unified experience for building, governing, operating, and consuming real-time data products. A shared user experience connects design and operations; supported runtimes execute the work across your chosen data environments.

Oracle Database replication configuration screen for schema mapping

Keep the systems that run your business in lock-step.

Capture committed changes with GoldenGate’s enterprise-proven change data capture (CDC) technology, preserve transaction context, and provide the highest levels of availability. It helps keep mission-critical systems synchronized and online through planned or unplanned disruptions along with zero downtime migrations.  Easy to use replication topology views enable you to understand how changes reach their intended targets and data consumers.

Transaction-aware CDC · Replication · Estate visibility

Data pipeline designer screen showing source, transform, and join nodes

Put operational change to work beyond the landing zone.

Combine ZeroETL, ETL, E-LT, Continuous Transformation and Loading (CTL), reverse ETL, and batch processing. Build Bronze, Silver, and Gold products using database-native or Apache Iceberg tables, with freshness matched to the business need.

ZeroETL · CTL · Medallion patterns

Diagram of dataset data product identity and consumer outputs

Give every product an identity that lasts.

Associate data with schema, endpoints, quality controls, processing evidence, pipeline-level lineage, and physical versions. Publish governed tables, streams, JSON products, and subscription APIs, and register supported tables in the catalog you choose.

Dataset lifecycle · Iceberg · Streaming and JSON

Infographic about dataset data quality rules and validation levels

Make data fitness visible before consumers act.

Activate data trust by design, not an afterthought. Automatically apply profiling, validation, quality checks, and configured security controls as data moves based on user-defined policies. Use Veridata to validate data with source-to-target comparison and repair; use transformation tests and quality checks for transformed data products. Build data trust by keeping the run evidence associated with each data product.

Validation · Veridata · Processing evidence

AI Pipeline Designer screen showing completed data join pipeline

Turn a good design into your team’s next starting point.

Build with natural-language assistance, a visual low-code editor, or a high-code domain-specific API. Review and test generated designs, then branch, merge, and deploy through controlled workflows. Reuse proven Datasets and patterns across teams.

AI-assisted design · Human review · Reusable patterns

Diagram of zone-by-zone engine placement across bronze, silver, and gold tiers

Put processing where the workload belongs.

Choose GoldenGate, Oracle AI Database, local Spark, Oracle AI Data Platform Spark, or supported external Spark for the work each engine supports. Keep data locality, performance, sovereignty, and total cost in the placement decision. Revalidate designs when changing engines.

Database pushdown · Spark · Data-local execution

Meet the Dataset

A data product with a lasting identity.

A Dataset is the durable logical lifecycle record for a governed data product. It keeps the product connected to its schema, controls, processing evidence, and physical versions as infrastructure changes beneath it. Teams can understand what they are consuming, how it was produced, and which controls apply—without tying the product’s identity to one database, catalog, or engine.

Infographic about governed dataset example for order state

Product Tour

Follow one order from commit to action.
A customer places an order. Inventory changes. Fulfillment needs to respond, and a demand-planning agent needs the current picture. Follow the data through four steps.

Begin where the business event becomes real.

The source database commits an order, an inventory decrement, and a shipment status. GoldenGate captures the change with transaction context, giving downstream processing an authoritative starting point. Illustrative journey. Source transaction semantics and supported capture paths depend on configuration.

Illustration of source transaction order status and fulfillment details

Prepare the product consumers need.

Apply mappings, quality rules, and the access controls for each product. Enrich with current fulfillment context, handle invalid records, and retain pipeline and run evidence. Veridata comparisons support eligible source-to-target copies. Transformed products also require transformation tests and quality checks.

Illustration of order product rules and validation checks

Capture once. Serve more than one need.

The same source capture can feed a synchronized operational copy, a conformed Iceberg table, an application stream, and a product interface for AI grounding. Each product has its own publication path and freshness requirements. Publication is not atomic across independent destinations or Iceberg tables.

Diagram of illustrative data products for one event and multiple outputs

Give the next decision a current view.

Fulfillment routes the order using current stock. A demand-planning agent receives updated operational context. An authorized reorder commits new business state, starting the next cycle. Applications and their owners retain decision authority, approvals, and action auditing. The fabric supplies the current, governed data.

Illustration showing order routed and AI demand agent notified
From the GoldenGate blog

New GoldenGate Data Fabric: trusted AI starts with the data that runs your business

By Jeff Pollock, Vice President of Product Management

AI agents can act on stale data faster—but not better. Oracle GoldenGate Data Fabric, the AI Truth Layer, turns committed changes across operational systems into governed, traceable, AI-ready data products, helping teams move from business events to confident action and choose where workloads run. GoldenGate Data Fabric is a governed, AI-ready data product layer as part of OCI GoldenGate on Oracle Cloud and multicloud platforms.

Read the blog

Latest GoldenGate Data Fabric resources

Frequently asked questions

  • Get to know GoldenGate Data Fabric.

    • What is Oracle GoldenGate Data Fabric?
    • How does GoldenGate Data Fabric extend the Oracle GoldenGate platform?
    • What is a Dataset in GoldenGate Data Fabric?
    • What does an operational truth layer mean for AI?
    • How does GoldenGate Data Fabric work with a lakehouse or Oracle AI Data Platform?
    • Which integration patterns and execution engines are supported?
    • Does it preserve transactions across every destination?
    • Which deployment and catalog options are in launch scope?
    • What capabilities are on the roadmap?
    • How should we measure the business value?

GET STARTED

Bring one business event. Build your next competitive edge.

 Choose one costly delay, one platform transition, or one AI use case that needs more current data. Define the product, agree the measures, and put the first outcome to work. Discuss a scoped plan-of-action with your Oracle team. Scope, timing, supported endpoints, and success measures are agreed for your environment.