---
updatedAt: 2026-06-23T11:50:34.000Z
agentTools:
  projectIndex: https://docs.phacetlabs.com/llms.txt
---

# Match

<br />

## Matching Column: Link records using matching

The **Matching column** lets you match input data from your current Table with reference data from another Table, automatically and at scale.

Use it when you want to reconcile data, enrich records, or validate that two datasets refer to the same item, entity, transaction, or reference..

## When to use Matching Columns

**Reconcile datasets:** Match internal records to master data sources like product catalogs, legal entities, or customer databases.

*Example: Reconcile payment logs with open invoices to confirm cleared transactions.*

**Validate business rules:** Detect inconsistencies between declared values and official references.

*Example: Check purchase prices against a negotiated supplier price list.*

**Standardize inputs:** Normalize naming conventions across systems by aligning to a trusted reference.

*Example: Match free-text supplier names to a legal entity register.*

## How It Works

Matching columns identify matches between input data from your current Table and reference data stored in another Table.

Two matching mechanisms are available:

* **Exact Match:** uses logic to identify identical matches between the input data and a reference dataset stored in another Table.
* **AI Match:** uses AI reasoning to identify matches that are not strictly identical between input data and a reference dataset stored in another Table.

Use the same Matching column type, then choose the matching tool that fits your data.

## AI Match vs Exact Match

Use **Exact Match** when values must match strictly.

Use **AI Match** when values are incomplete, inconsistent, or semantically close but not identical.

| Use Exact Match when…                | Use AI Match when…                                 |
| ------------------------------------ | -------------------------------------------------- |
| You have a stable unique identifier. | You do not have a reliable shared key.             |
| Values must be identical.            | Values can be written differently.                 |
| The match should be deterministic.   | The match requires interpretation.                 |
| Example: `SKU = SKU`.                | Example: “Tomatoes 5kg box” ≈ “Tomate colis 5 kg”. |

Examples:

* Use **Exact Match** to match invoice lines with a price list using a SKU.
* Use **AI Match** to match invoice lines with a price list using product designation, packaging, and unit.
* Use **Exact Match** to match suppliers by SIRET.
* Use **AI Match** to match suppliers by name, address, and email domain.

<Callout icon="💡" theme="default">
  You can combine both approaches in AI Match instructions.

  For example, instruct AI Match that a SKU must match exactly when available, then use AI reasoning on designation and packaging when the SKU is missing or unreliable.
</Callout>

## How to set up a Matching column

To set up matching, open one of the detailed implementation articles:

* [Set up Exact Match](https://docs.phacetlabs.com/docs/exact-match)
* [Set up AI Match](https://docs.phacetlabs.com/docs/ai-match)

<br />

<br />

<br />

<br />

<br />

<br />