AI Contact Reason tagging is a large-language-model approach to understanding what happened in each interaction and automatically applying labels to classify them by intent. With AI Contact Reasons, every interaction is assigned a tag from a predefined list. This happens automatically and consistently. It’s designed to give you well-defined intent categories that are easier to explore and analyse to get the insights you need.

Why we built this

This feature was initially developed to provide a structured, automated way to categorise entire interactions without the need for manual selection or complex trigger setups.

Until now, this has been implemented through collaboration with EdgeTier’s team: you define your categorisation goals, and we create specialised tagging models to help you achieve them. You could always see these tags on your interactions across EdgeTier, but you could not visualise them or their underlying rules in any dedicated page.

We’ve now added an interface to allow any EdgeTier user in your organisation to view the Contact Reason categories being applied, so it’s easier to understand and trust the AI output.

This new page is a home for your Contact Reasons, and a first step to full self-service AI tagging in the interface. It allows you to view all categories and the definitions and use cases behind them.

Click on any tag to see how it’s configured, and select the Performance tab to see which interactions it’s been applied to recently. The Explore button brings you to all tagged conversations in the Explore screen for further analysis.

Note: Undefined Topics also has its own view within Contact Reasons, so you can quickly see if your existing categories are giving you the coverage you need, or if a large volume of interactions are yet undefined.

Other Changes

As part of this release, we’ve created a new ‘Tags’ section in the main EdgeTier navigation. The existing ‘Tags and Triggers’ page has moved here, creating a centralised hub for all tag management.

Coming soon: Tweak your setup and add new categories and definitions in the interface.