Introduction

The Memory tab in wolkvox Agent allows agents to review historical and contextual customer information automatically generated by Artificial Intelligence during customer interactions. This functionality integrates with wolkvox's conversation analytics capabilities to consolidate metrics, customer attributes, and recent interaction outcomes into a single view.

This allows agents to quickly understand previous customer experiences, sentiment, topics discussed, and outcomes achieved without manually reviewing their entire conversation history.

For Memory to generate customer information, there must be at least 6 prior chat conversations associated with the customer. These conversations can come from channels such as a web chat widget created in wolkvox Manager, WhatsApp, Facebook Messenger, or Instagram Direct.

Once available, Memory uses the customer's last 10 processed interactions as a reference. These can include chats and calls, meaning the displayed information updates as new interactions occur and does not represent an unlimited history.

How to use the Memory tab in wolkvox Agent?

1. Access voice interactions

1. From the left side menu in wolkvox Agent, go to the “Voice” section.
2. During interactions, next to the “Call Info” tab, you will find the “Memory” tab. Click on it to view available customer information.
3. Information available in Memory Agent:
   • At the top of Memory, a customer identification card is displayed, allowing you to see information such as:
     • Customer ID or associated number.
     • Contact channel.
     • Total number of recorded interactions.
     • Date and time of last contact.
   • Below, you will find several metrics calculated from recent interactions. These metrics are calculated using the last 10 available interactions as a reference: 
     • Avg Sentiment: Displays the average sentiment identified in recent customer conversations and its corresponding classification, allowing you to quickly determine if their experience has been mostly positive, neutral, or negative.
     • Avg CSAT: Displays the customer's average CSAT score on a 5-point scale, accompanied by a visual classification of the result.
     • FCR Rate: Indicates the percentage of interactions considered resolved on first contact. It also shows how many analyzed interactions were identified as resolved.
     • NPS Mix: Summarizes customer behavior according to NPS classification, allowing you to visualize the proportion associated with categories such as promoter, passive, or detractor.
     • Channel Distribution: Allows you to identify what proportion of recent interactions occurred across each channel. For example, it can show the share of voice calls compared to other channels used by the customer. 

Scrolling down within the Memory tab reveals additional information generated from the customer's history. 

• Customer Profile: Displays data associated with the contact, including:
  • Phone
  • Channel type
  • Source channel
• Memory Summary: Displays AI-generated text that synthesizes relevant information identified across customer conversations. This summary may include details such as prior contact context, observed behavior, recurring needs, topics discussed, and other useful information so agents can quickly understand customer history before or during a new interaction.
• Likes: Displays interests, preferences, or relevant items identified from previous customer conversations.
• Personal Notes: Consolidates additional contextual information gathered from interactions to help agents provide more continuous service. Information in these sections is generated automatically from available data in analyzed conversations.

At the bottom, you will find the Recent Interactions section, which displays up to the last 10 interactions associated with the customer.

Each record may present information such as:

• Interaction date and time.
• Channel, for example Chat-WhatsApp or Voice.
• Duration.
• Sentiment and its score.
• CSAT result.
• FCR result.
• NPS classification.
• Main topic identified.
• Summary of what occurred during the conversation.
• Identification of the agent who handled the service.

This view allows agents to quickly review past customer experiences and detect recurring issues, unresolved contacts, or changes in perception over time.

As new conversations are logged, Memory will continue using the most recent interactions as a reference, maintaining an analysis window of up to 10 interactions to provide agents with updated, relevant information during customer service.