This translates into a powerful cockpit that can be used to manage and improve the system performances, making it possible to measure benefits and detect improvement areas. The improvement and evolution of dialogue flows are crucial aspects when managing a conversational engine since they are strictly tied to the quality of the service provided.
The Actionable Reporting Dashboard offers tools that allow you to have a quick and efficient overview of the functioning of the chatbot and at the same time a sufficient
detail to get an accurate idea of any changes or improvements
The New Charts Dashboard groups into four distinct panels:
Quantitative KPIs trace:
Qualitative KPIs allow the analysis of elements such as:
POSSIBILITY OF CUSTOMIZED REPORTS
In addition to creating the predefined reports described so far, the Convy Admin the interface allows the creation of customized reports and graphs based on the data
collected by the system.
All reports produced can be exported in Excel format.
Moreover, since the information gathered by the reporting system is stored in a relational database, it is possible to prepare export scripts to automatically transfer
the collected data to external Business Intelligence systems or analysis.
Starting from the selected time interval (Created AT – CLosed At) this funcionality allows to have a complete overview of all the contacts for each customer under ‘User ID’, of the specific Dialog Flow that has been followed and of the ‘Channel’ used for the conversation with the ChatBOT (Facebook, Telegram, Twitter…)
By clicking on View Detail it is possible to see in a greater detail the conversation for each customer (Gender, Timezone…) and a complete History of the Chat
Convy Admin allows to have statistics related to intent recognition: how often the ChatBOT recognized a specific user intent, with the related degree of confidence
For instance, let’s consider the intent “Sale”: the integrated semantic engine recognized this user intent 100% of the times, 36 out of 36.
The 8% refers to the total of all intents considered in the specific time interval: in this way it is possible to build a ‘ranking’ of the topics that customers asked more frequently to the ChatBOT in a specific time interval.
The detail [+] allows to look into a specific semantic field and to see the ‘Path’ that led to that specific intent recognition: in this way it is possible to highlight the criticalities and improvements, and to refine the semantic interpretation
This feature allows to select a specific conversation flow, in order to obtain specific information in the selected time interval
For the selected flow Convy Admin shows the total number of incoming conversations, how many are unique and how many terminated or handed over to a human operator: this allows to understand how users navigate through the conversation flows
Changing the perspective, this feature allows to select a specific intent of the semantic engine and to analyze the degree of confidence by which the engine recognizes sentences and request involving the intent
The above picture displays the number of occurrences based on the confidence range of the semantic engine; this intent has been recognized 41 times with a degree of confidence ranging from 85 to 100% and only one time in the range from 0 to 40% with a semantic confidence score of 20.09%
©2019 Eudata | HQ Via dei Valtorta 48, 20127, Milano | Branch Office Viale Vittorio Veneto, 8/D, 21013 Gallarate (VA) | P.IVA 12421000154 | TEL. +39 0245495098