Learn

What is Reconciliation?

Reconciliation is the process of grouping related Artifacts together to form Assets. It is the bridge between raw discovery, where you find individual pieces of AI evidence, and structured governance, where you manage complete AI systems.

Why is Reconciliation needed?

When the platform scans your connected Sources, it discovers Artifacts independently. A training script in a code repository doesn't automatically know it belongs to the same AI system as a deployed endpoint in your cloud environment or a project document in your collaboration platform. Reconciliation is the step that makes these connections, linking related Artifacts into a single, unified Asset.

Without Reconciliation, your governance team would be looking at a long list of disconnected evidence items with no clear picture of which items belong to which AI system. Reconciliation turns that list into an organized portfolio of AI systems.

How Reconciliation works

1. The platform discovers Artifacts across your connected Sources through scanning

2. It analyzes relationships between Artifacts based on metadata, naming patterns, ownership, and other contextual signals

3. Related Artifacts are proposed as groupings that could form a single Asset

4. Your governance team reviews these proposed groupings and confirms, adjusts, or rejects them

5. Confirmed groupings become Assets in your AI inventory

The role of your governance team

Reconciliation is not a fully automated, hands off process. The platform does the heavy lifting of analyzing relationships and proposing groupings, but your team retains control. You can:

  • Review proposed groupings before they become Assets
  • Merge Artifacts that the platform didn't automatically connect
  • Split groupings that incorrectly combine unrelated Artifacts
  • Manually assign orphaned Artifacts to existing Assets
  • Create new Assets from Artifacts that don't fit into any existing grouping

This balance of automation and human oversight ensures that your AI inventory is both comprehensive and accurate.

What Reconciliation looks like in practice

Before Reconciliation, you have a flat list of Artifacts: scattered evidence across multiple Sources with no clear organization. After Reconciliation, you have a structured inventory of Assets: clearly defined AI systems, each with all of its related Artifacts linked together, ready for governance.

Before - Dozens or hundreds of disconnected Artifacts across different Sources with no clear relationships

After - A clean, organized inventory where each Asset represents one AI system with all its related evidence connected

Why Holistic AI

Reconciliation is one of the most overlooked steps in AI governance. Many platforms either skip it entirely, forcing teams to manually organize everything, or fully automate it with no room for human review. The Holistic AI Governance Platform strikes the right balance. It uses intelligent analysis to propose groupings while keeping your team in control of the final decisions. The result is an AI inventory that is both complete and trustworthy.

See how it works  |  Book a demo

Holistic AI Governance Platform  |  Learn Section  |  AI Discovery & Inventory

Share this

See Holistic AI Governance Platform in action

See how Holistic AI puts these concepts into practice.
Request a Demo

Stay informed with the Latest News & Updates