Skip to main content
AI Data Engine

What's new in AI Data Engine

Contributors netapp-dbagwell RSS
PDFs

Learn about the latest capabilities available in AI Data Engine (AIDE) releases for AI data management, including metadata cataloging, workspace management, and flexible deployment options.

What's new in AIDE 1.0.0

AIDE 1.0.0 (released April 30, 2026) introduces AIDE as a software-only deployment on third-party servers. This deployment option allows you to run AIDE software on your own RHEL 9.7 servers without requiring NetApp Data Compute Nodes (DCNs). For this release, AIDE software on third-party servers supports metadata-focused capabilities.

Customer-managed hardware deployment option

The AIDE software on third-party servers deployment option provides flexibility for you to use your own hardware with AIDE's metadata management capabilities while managing the underlying hardware and OS yourself. Key characteristics of this deployment model include:

  • Three-node homogeneous clusters required

  • You install OS and manage OS security patching

  • NetApp provides AIDE software stack and support

  • Requires ONTAP 9.18.1 or later

  • Requires an ONTAP AFX storage system

AIDE Metadata Engine basic support

AIDE software on third-party servers supports the core Metadata Engine features of AIDE, enabling you to catalog and organize your metadata for AI workloads.

For this release, AIDE software on third-party servers does not support GPU-dependent services such as vectorization, OCR, and enrichment, and it does not support data collections, vector embeddings, or RAG endpoints for semantic search. An upgrade path to full AIDE capabilities is supported when those capabilities become available.

Metadata Engine features supported

AIDE software on third-party servers provides the following metadata capabilities:

  • Workspace creation and management

  • Automated metadata extraction and cataloging

  • Metadata search and filtering using REST APIs

  • Data Sync for automated data currency

  • Metadata export functionality

Metadata Engine search and browsing

In AIDE 1.0.0, Metadata Engine functionality is available in AI Data Engine Console. You can search, browse, filter, and export metadata from a centralized view across all workspaces. You can explore metadata from a new page in AIDE Console that's included with all AIDE software on third-party servers deployments. This page can be used after workspace creation and metadata extraction are complete.

What's new in AIDE 9.18.1U0/U1 initial release

AI Data Engine (AIDE) 9.18.1 U0/U1 is the initial release of NetApp's platform for AI data management. This release introduces a metadata engine and management workflows that enable organizations to catalog and organize unstructured data for AI workloads, providing the foundation for advanced governance and vectorization capabilities. Advanced governance (guardrails) and vectorization are available for customers who have the appropriate AI Data Engine licenses.

AIDE 9.18.1U0 introduces the following foundational capabilities:

Metadata Engine for AI data cataloging

The initial release includes a metadata engine that catalogs files and objects across ONTAP clusters.

Key features include:

  • Automated extraction of metadata (core and extended attributes, object tags) from local and remote ONTAP volumes on peered clusters.

  • Centralized querying and filtering REST APIs for applications requiring a global view of enterprise data.

  • Scalable metadata storage.

  • Automatic metadata extraction triggered during workspace creation.

Workspace management

Workspaces provide logical grouping of data sources (volumes) for AI projects.

The initial release supports:

  • Creation of workspaces spanning local and remote ONTAP volumes (using cluster peering).

  • Assignment of access controls to workspaces, supporting multi-user and multi-tenant environments.

  • Automatic metadata extraction and catalog population upon workspace creation.

Data Sync for automated data currency

Data Sync keeps metadata catalogs and data collections current as source data changes, without manual intervention.

Key features include:

  • Automated synchronization of data from remote or local ONTAP clusters using policy-driven SnapMirror replication.

  • Incremental updates that propagate only modified data, reducing overhead.

  • Configurable refresh intervals per workspace.

  • Workspace-level monitoring of sync status and activity.

Cluster setup and management

The initial release includes the following workflows:

  • Discovery and addition of data compute nodes (DCNs) during cluster setup.

  • Creation of dedicated metadata storage VMs for the metadata engine.

  • Configuration of Data Engine service interfaces for cluster-wide metadata access.

  • Peering with other ONTAP clusters to extend metadata cataloging across the data estate.

OpenID Connect (OIDC) authentication

  • OIDC/OAuth-based authentication for secure access to ONTAP System Manager and Data Engine Console with Microsoft Entra ID and Active Directory Federation Services (ADFS).

  • Role-based access controls for workspace and metadata management.

Advanced data curation and governance capabilities

The following capabilities are available with the AIDE 9.18.1U1 update for customers who have the appropriate AIDE licenses:

  • Vectorization and RAG: Creation of data collections, embeddings, and retrieval endpoints in AI Data Engine Console, using metadata from AIDE workspaces.

  • Guardrail-based governance: Definition of guardrail policies in AI Data Engine Console and association of those policies with workspaces in ONTAP System Manager.

Supported hardware and platforms

AIDE 9.18.1 U0 and U1 run on ONTAP AI data platform clusters that combine:

  • AFX 1K storage nodes

  • NetApp data compute nodes

Related information