Amber Chat Bot

Client Mobile Web

Amber Chat Bot

Let’s be honest—calling customer service can be a frustrating and exhausting experience for anyone…introducing Amber!

___

Concept

Amber is Ambetter’s (CENTENE Corporation) chatbot virtual assistant.  Amber has become a Jr. Member Services Associate with the task of guiding members through the complicated world of health insurance. She can help with things like;

1. Help members understand the basics of their plan (e.g., deductible, premium costs, etc.
2. Help members make a premium payment
3. Explain certain medical terminology, such as “pre-authorization
4. Troubleshoot issues such as finding appropriate tax documents
5. Provide plan-related addresses and contact information
6. Help members find a medical provider
7. Search coverage information

____

Opportunity

Conversational AI is a powerful tool for automating customer interactions. To deliver a worthwhile customer experience, it’s important to look beyond the technical aspects of the technology and also pay attention to how customers will interact with your virtual agent.

• You need to study, test, design, and account for;
• How a virtual agent can influence customer behavior
• Recommendations for how to navigate natural conversation barriers
• What is necessary to improve longer conversation flows
• The importance of a virtual agent’s pronouns
• The secret weapon of interaction design – empathy in tone and voice

__

That last bullet is what drives my passion for design. Design is not just what looks good or even the ease of use; design powerfully influences emotion and connection. When you walk away from an experience, how does it make you feel? Do you feel like you would come back to that experience? Do you trust that experience?

So when creating a customer service chatbot for health care, it is not creating just any chatbot. Healthcare is a sensitive subject. We need to be clear and informative while also being kind and empathetic.

___

The Impact

I collaborated with the Conversational Design Team to create new interactions like Coverage Search, where users can ask Amber what a specific medical service might cost. Amber recognizes both broad and specific search terms and delivers personalized results directly within the chat interface.

Building this feature presented several challenges. Training a bot to not only understand intent but also surface accurate, contextual results required extensive iteration. For example, how do we display multiple relevant costs—like ambulance, facility, and physician fees for an ER visit—without overwhelming the user? Or how do we explain that a single procedure, such as a colonoscopy, may be covered differently depending on whether it’s preventative or diagnostic?

To address these challenges, we designed multiple conversation scenarios and conducted both moderated and unmoderated usability tests to observe user reactions and gather feedback. The result was a more intelligent, user-centered experience that helps members quickly and confidently find the cost information they need.