The Kore.ai XO Platform provides multiple dashboards to analyze real-time data. The data is generated as the virtual assistant interacts with the users. This analysis is helpful to make better decisions and improves the business value.
In this section, we use the Magento store assist bot as an example to view the various analytics dashboard. Because the Magento store assist bot has both web/mobile client and Twilio voice enabled, it provides information for both these channels in the dashboards.
The Overview dashboard provides a summary of various dashboards such as the Conversations Dashboard, Users Dashboard, and Performance Dashboard. Under the overview of the Conversations Dashboard in the following figure, we see how many conversation sessions have been completed along with the drop offs (if any) and total number of messages. The filter bar at the top provides a customizable date range to view the analytics data. The overview of the Users Dashboard provides information related to the number of unique users. The overview of the Performance Dashboard provides information such as if the intent is identified correctly, the API execution, script execution rate, and so on.
Figure 8. Overview dashboard
This page provides information about how conversation sessions are managed and handled. It primarily covers whether calls were answered by the virtual assistant, passed on to agents, or are incomplete. You can also see various other metrics, such as the trend in conversation sessions, analysis of the conversation path, distribution of sessions by channel, virtual assistant engagement, and the breakdown of self-service, drop-off, and agent transfer sessions.
Figure 9. Dashboard showing conversation sessions and trends
You can also look at the sessions used in web and voice channels, the conversation path (for developers to understand what the user query was and how the virtual assistant determined the tasks) and engagement analysis.
This page serves as a hub for data on user trends and returning user trends. It also includes insights such as the total number of unique users, the number of returning users, the number of new users, and weekly or daily retention rates for different user groups.
Figure 10. Users Dashboard
This page provides insights into the performance of a virtual assistant. A virtual assistant designer can see how many intents were identified, how many tasks were completed using the virtual assistant, and how many services or scripts failed during interactions. The performance dashboard offers all this crucial information.
Figure 11. Performance Dashboard