Digital Contact Centre Solution for Customer activities generative summary and sentiment analysis
Description
In a housing industry, a contact who could a tenant may have raised repair complaints which comes through different channels like emails, phone, SMS etc. Even in some cases tenants would have raised concerns on social media platforms like facebook, instragram, twitter. All these interactions on a contact profile gets registered in the timeline of the customer service hub in a Dynamics 365 CE application
The solution analyses this customer timeline data, pushes them into a real time event stream in Fabric and through pipelines pushes everything to a delta lake table in lakehouse.From lakehouse it runs through Azure Open AI models and group by all interactions and finally responds with a summary of what happening on the contact profile.
Also it returns with the with the sentiment analysis and the localized translation of the customer summarized activities data.All of these get registered on the customer profile in the Dynamics application
By doing all these, it helps businesses to prioritize their customers and devise different strategies for different customer segment based on the sentiment score
It helps Customer service managers to prioritize their more dissatified customers for quick followup, upsell/cross-sell for highly satisfied customers. They can just understand what is happening on the customer looking at the feedback summary and not going through hundreds of timeline information
The wiki page in the below github project documents the whole solution approach
Project name
Digital Contact Centre Solution for Customer activities generative summary and sentiment analysis
Description
In a housing industry, a contact who could a tenant may have raised repair complaints which comes through different channels like emails, phone, SMS etc. Even in some cases tenants would have raised concerns on social media platforms like facebook, instragram, twitter. All these interactions on a contact profile gets registered in the timeline of the customer service hub in a Dynamics 365 CE application
The solution analyses this customer timeline data, pushes them into a real time event stream in Fabric and through pipelines pushes everything to a delta lake table in lakehouse.From lakehouse it runs through Azure Open AI models and group by all interactions and finally responds with a summary of what happening on the contact profile.
Also it returns with the with the sentiment analysis and the localized translation of the customer summarized activities data.All of these get registered on the customer profile in the Dynamics application
By doing all these, it helps businesses to prioritize their customers and devise different strategies for different customer segment based on the sentiment score It helps Customer service managers to prioritize their more dissatified customers for quick followup, upsell/cross-sell for highly satisfied customers. They can just understand what is happening on the customer looking at the feedback summary and not going through hundreds of timeline information
The wiki page in the below github project documents the whole solution approach
Project Repository URL
https://github.com/Sapgithub/Sapgithub
Project video
https://1drv.ms/v/c/c89ad874e90247fe/EcQIJbWHkpZJsmnQ17E6GSkBfhAVWKywFOTrOmoIIYXhnQ?e=F10B29
Team members
Sapgithub