<aside> 💡

Alongside other interaction analytics, our platform provides you with resolution tags.

</aside>

How it works

Emotion and Sentiment Detection

EdgeTier analyses every message and utterance from customers for emotion and sentiment. As part of this, our platform detects an interaction’s resolution status. This LLM-powered feature can identify whether a customer’s query was resolved based on signals within the conversation.

Resolution status is a filter within EdgeTier. You can filter for resolved or unresolved conversations, providing another powerful data-split to manage agent quality and identify problematic interactions or topics.

<aside> ❗

EdgeTier already stores an ‘Survey Response - Resolved’ (originally ‘Is query resolved’) metric for some customers. This is a True/False survey response metric, and not related to resolution status.

</aside>

Our model uses an AI-powered "resolution" tag that helps our clients see how well their support interactions are being handled. There are four main states for interactions:

• Resolved - This indicates that the interaction query was resolved by the agent
• Unresolved - This can indicate that the interaction query was not resolved by the agent before the interaction was closed or that the interaction query is still ongoing.
• No resolution needed - This indicates that while no resolution was detected explicitly, the service determined that the query did not require a resolution.
• N/A - Interaction was not processed through the AI Resolution service.

Resolution can provide insights into how well agents are able to answer queries on certain topics, as well as assisting with agent QA for specific issues.