Homepage


About Me

Projects


AIsurance: AI Agent for Insurance

Main Role: UX/UI Design






Explore Gemini AI: 


Our team explored in the Google Cloud Vertex AI Agent Builder Hackathon that Google's Gemini is 

  • better at handling multimodal inputs than other AI Agents,
  • has a gentle and friendly tone of voice, which is suitable for service.


Based on that, we believe that rapid summarization and reporting of complex customer service information in SaaS services will greatly improve the comprehension efficiency of each participant.




The Problem:
Inefficient Customer Service Message Processing in Insurance Companies



Our team chose insurance claims as a case study. Due to the complexity of the process and the amount of paperwork, insurance companies need to spend a lot of time and money on claims documentation. In addition, manual data entry is prone to errors that can lead to fraud and other problems. 

As a result, we can leverage Gemini to help insurers streamline the claims process, process users' multimodal information in real-time, including video, voice, text, etc., and merge the process of training AI Agents into their daily workflow.







User Research



Based on actual interviews and available insurance company information, we have summarized the key points of the current vehicle insurance claims process as follows:

Customers

      ·  Need to be able to contact insurance companies anywhere, anytime.
      ·  Need to provide evidence in the most efficient and accurate way

Insurance Claim Adjusters:

      ·  Need to provide personalized answers for each client
      ·  Requires cumulative experience for actual case handling





Design Goals



1.  24/7 Claims Assistance for Customers:Provide users with clear and organized guidance steps to quickly collect and analyze claims information;

2.  Automated Claims Report and Progress Updates:Help all staff quickly follow up on news updates and provide prompts for next steps or disputes based on policy;

3.  Integrate Chatbot Training into Daily Workflow:Typical cases can be entered into the chatbot's model library for reference in subsequent related cases.





User Pain Points



Based on the user journeys I created in collaboration with the product manager, we summarized the pain points of our users as:

Customers:

  • Allow customers to provide evidence in their expression and language


Insurance Claim Adjusters:

  • AI helps insurance adjusters quickly source policies
  • Claim adjusters train and develop AI Agent as an intern





User Flow






Wireframe




Final Showcase















Takeaways



1.  Requirement of close collaboration and meetings with product managersTight collaboration between designers and product managers enables rapid decision-making in various directions and low-cost experimentation during the early stages of product development.

2.  Use bullet points to clearly articulate the focus of each stepClear articulation of the key points of the steps can help all participants understand complex content, such as customer service, legal, front-end, and data analysis.

3.  AI as a long-term Intern!Conceive of AI as a long-term intern assistant in the workplace and explore mechanisms to foster its growth.



Project 1 Custom-made MouseStartup Launching
Project 2 AIsurance AI Agent for Insurance
Project 3 Package DepositBlockchain for Sustainability




Welcome to contact me:       yuhanghan97@gmail.com