Discover how an insurance company with more than 120 years of experience managed to modernize its operations by incorporating AI.
Problem
The telephone service executives had to use four different systems (Salesforce, insurance core, legacy system, and document search) to respond to customers, which was time-consuming and negatively impacted the customer experience.
Objective
Consolidate all tools into an AI assistant that unifies and simplifies the information search process, helping customer service executives provide faster and more efficient telephone support.
Challenges
The biggest challenge was deploying on Google Cloud Platform (GCP) and integrating the APIs due to their complexity and the use of OpenAPI 3.0. Additionally, it was necessary to manage authentication with AWS Sign instead of OpenAPI.
Solution
We worked with the client's Generative AI division to develop an AI assistant that simplifies and streamlines the service provided by customer care agents, which integrates the multiple systems the agents must consult to obtain information. This implementation allows users to ask a question in natural language to the AI assistant, which then analyzes it and selects the best tool to provide the answer, whether through an API or a RAG (Retrieval- Augmented Generation). In addition, the AI assistant formats the resulting answer and presents it in a way that is clear and understandable to the user, making sense of the information under a single criteria.
Technically, the solution consisted of creating a web server that was contained and deployed in the cloud using CloudRun in GCP. his web server was implemented as a provider model by GeneXus and held the description of all the tools available to answer user questions.
The solution, called Assistant of Assistants, currently features an API assistant that responds to information contained in the CRM, insurance core, and legacy system, in addition to three RAG assistants that respond to a base of more than 100 documents.
Developing this AI solution with Globant Enterprise AI ensured:
Results
The solution is currently in beta and undergoing UAT testing. However, the implementation of this AI assistant is expected to democratize knowledge among customer service agents, allowing them to focus on solving complex problems rather than repetitive tasks. In addition, this Assistant of Assistants developed with Globant Enterprise AI has sparked new conversations with the client about additional use cases that will be explored in the future.
The telephone service executives had to use four different systems (Salesforce, insurance core, legacy system, and document search) to respond to customers, which was time-consuming and negatively impacted the customer experience.
Objective
Consolidate all tools into an AI assistant that unifies and simplifies the information search process, helping customer service executives provide faster and more efficient telephone support.
Challenges
The biggest challenge was deploying on Google Cloud Platform (GCP) and integrating the APIs due to their complexity and the use of OpenAPI 3.0. Additionally, it was necessary to manage authentication with AWS Sign instead of OpenAPI.
Solution
We worked with the client's Generative AI division to develop an AI assistant that simplifies and streamlines the service provided by customer care agents, which integrates the multiple systems the agents must consult to obtain information. This implementation allows users to ask a question in natural language to the AI assistant, which then analyzes it and selects the best tool to provide the answer, whether through an API or a RAG (Retrieval- Augmented Generation). In addition, the AI assistant formats the resulting answer and presents it in a way that is clear and understandable to the user, making sense of the information under a single criteria.
Technically, the solution consisted of creating a web server that was contained and deployed in the cloud using CloudRun in GCP. his web server was implemented as a provider model by GeneXus and held the description of all the tools available to answer user questions.
The solution, called Assistant of Assistants, currently features an API assistant that responds to information contained in the CRM, insurance core, and legacy system, in addition to three RAG assistants that respond to a base of more than 100 documents.
Developing this AI solution with Globant Enterprise AI ensured:
• A significant acceleration of implementation times.
• Data security and privacy in handling and processing.
• Efficiency and accuracy in processing vast
amounts of documentation without
compromising performance.
Results
The solution is currently in beta and undergoing UAT testing. However, the implementation of this AI assistant is expected to democratize knowledge among customer service agents, allowing them to focus on solving complex problems rather than repetitive tasks. In addition, this Assistant of Assistants developed with Globant Enterprise AI has sparked new conversations with the client about additional use cases that will be explored in the future.