IBM GenAI Connector for Unstructured Data

The presented connector is configured to improve your Customer Service Process. WatsonX is processing written service requests to create categories such as “reason for contact”, “Root Causes”. This data is then infused into the Celonis EMS to analyze and improve ticketing times, customer satisfaction and more. On top of that, audio transcript data can be processed by leveraging the NLP capabilities of WatsonX.

IBM WatsonX GenAI is a pre-trained AI which pre-processes the unstructured data, summarizing and categorizing it - depending on your needs - before the data is fed into the Celonis EMS.

For the Customer Service “reason for contact”, “Root Causes”, “topic”, “customer behavior”, “request summary” was created from the original request and fed into the analysis.

Conditions for the usage of the IBM GenAI Connector for Unstructured Data are the joint setup of the data interfaces, the adaption of prompts for the GenAI model (if needed), the integration into the existing data model in Celonis as well as the installation of the included analysis. If the Use Case differs from Customer Service, new prompts & analyses must be defined additionally.

The solution follows a tool agnostic approach - The included architecture template can be leveraged for any GenAI solution and can be customized to work with leading GenAI platforms like WatsonX, Azure/OpenAI, Amazon Bedrock and others. The agnostic approach provides you with flexibility, cost-effectiveness and adaptability, allowing you to choose the best solutions for specific tasks without being tied to a particular vendor or technology. This approach encourages innovation, improves collaboration, and helps future-proof the organization against rapid technological advancements.