Recipes Overview

Recipes are guided, repeatable processes you set up once and return to whenever your data needs attention. Each recipe carries one data quality job from start to finish, in phases and steps, and produces ordinary platform objects along the way: quality checks, scans, anomalies, and outputs in your enrichment destination. Nothing a recipe creates is a new kind of object, so everything it produces can be managed from the pages you already use.

Open Recipes from the left side panel to see the recipes available. Each card describes what the recipe does and lists its main stages. Click Get Started to begin.

Every recipe keeps your progress so you can leave and come back, but where that progress lives differs. Entity Resolution and Data Reconciliation keep it in your browser session, so their card shows an In progress pill and its button reads Continue. Data Suitability keeps it on each AI Use Case instead, so its card always reads Get Started; resume from the record list the recipe opens on, or from the AI Use Case page.

When AgentQ is configured, recipes offer AI suggestions and summaries at defined points, such as recommending fields or interpreting results. Every step also works without them.

This section covers:

• Entity Resolution: takes a profiled table from suspected duplicates to a deduplicated golden set written to your enrichment destination.
• Data Reconciliation: compares an asset with a reference asset row by row, lets you decide which side is right for every mismatch, and writes the keep and remove sets to your enrichment destination.
• Data Suitability: turns the data feeding an AI model into an AI Use Case, a governance record with a certified, point-in-time evidence record an auditor can check.

Entity Resolution

Data Reconciliation

Data Suitability