What is an Artifact?

An Artifact is the smallest unit of discovery in the Holistic AI Governance Platform. It represents a single, raw piece of evidence that was found during a scan and classified as AI related. Artifacts are the building blocks of your AI inventory.

An Artifact is an atomic record, meaning it represents one specific item from one specific Source. It could be a model training script found in a code repository, a deployed endpoint discovered in a cloud environment, or a document about an AI project found in a collaboration tool. Each Artifact stands on its own as a discrete piece of evidence that AI activity exists somewhere in your organization.

• Source - Where it was found, which connected platform and specific location within that platform
• Classification - Whether it was identified as AI or Non AI during the classification process
• Metadata - Information about the item such as its name, file type, location, date discovered, last modified date, and associated owner or team
• Evidence - The raw details that support why this item was classified as AI related
• Status - Whether it has been reviewed, is pending review, or has been grouped into an Asset

How Artifacts are created

1. A Source is connected to the platform

2. A Scan runs against that Source

3. The platform discovers content and classifies it

4. Items classified as AI related are created as Artifacts

5. Each Artifact is stored with its full metadata and classification details

Artifacts are created automatically. Your governance team does not need to manually log or register them. The platform handles the entire process from discovery through classification to Artifact creation.

How Artifacts fit into the bigger picture

Artifacts on their own are raw data points. A single AI system might generate multiple Artifacts across different Sources. For instance, a machine learning model might have training code in one repository, deployment configuration in a cloud environment, and documentation in a collaboration platform. Each of these would be a separate Artifact.

The next step in the governance journey is grouping related Artifacts together into Assets, which represent complete AI systems. This grouping process is called Reconciliation.

Managing Artifacts

The platform provides a dedicated Artifacts management view where your governance team can browse, search, filter, and review all discovered Artifacts. You can filter by Source, classification, date, status, and more. This gives you granular visibility into the raw evidence behind your AI inventory.