Co-browsing Service with Sensitive Data Masking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Customer service agents face challenges in providing assistance during web site interactions, as they cannot visually see the customer's view, which hampers their ability to effectively troubleshoot issues without accessing sensitive data.
Innovation Solution
A co-browse service that uses ECMAScript, such as JavaScript, to mirror browser state from the visitor to the agent, masking sensitive data and allowing the agent to view the web page without requiring local state information, enabling synchronized views while maintaining privacy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the agent is provided with the customer's browser state to see what the customer is viewing, then the agent's ability to provide assistance is improved, but the risk of accessing sensitive customer data increases
Solution Approach 1:
The patent segments the browser state into two distinct parts: visible state (DOM structure, layout, non-sensitive elements) and sensitive state (form field values, personal data). The co-browsing system transmits only the visible state to the agent, while sensitive data remains isolated in the customer's browser. This segmentation allows the agent to see what the customer sees without exposing sensitive information.
Solution Approach 2:
The patent introduces a co-browsing service as an intermediary between the customer's browser and the agent's view. This service acts as a mediator that receives the complete browser state from the customer, filters out sensitive information according to predefined masking rules, and transmits only the appropriate visible state to the agent. The intermediary ensures that sensitive data never directly reaches the agent's interface.
2Measurement precision
If the agent's browser runs the same script as the visitor's browser to synchronize views, then the view synchronization is improved, but the device complexity and security risks increase
Solution Approach 1:
The patent creates a simplified copy of the browser state for transmission to the agent. Instead of transmitting the complete browser state including sensitive data and complex runtime information, the system generates a masked copy containing only the visible DOM structure and relevant state information. This copying approach maintains view synchronization while reducing complexity and security risks.
Solution Approach 2:
The patent extracts only the necessary visible state information from the complete browser state for transmission to the agent. The system separates and extracts DOM structure, layout information, and non-sensitive elements, while leaving out sensitive form data and unnecessary browser runtime state. This extraction reduces the amount of data transmitted and simplifies the agent's browser state management.
3Reliability
If sensitive data is masked from the agent's view, then customer privacy is improved, but the agent's ability to troubleshoot certain issues may be worsened
Solution Approach 1:
The patent applies different quality levels to different parts of the browser state. Sensitive elements such as form fields containing personal information are completely masked with no visibility to the agent. Non-sensitive elements such as navigation structure, buttons, and public information remain fully visible. This local quality differentiation ensures privacy protection where needed while maintaining troubleshooting capability where appropriate.
Solution Approach 2:
The patent implements a feedback mechanism where the agent can request additional information or clarification from the customer about masked areas. When the agent encounters a troubleshooting issue that requires information from a masked sensitive field, the system provides a feedback channel for the agent to ask the customer to verbally provide the necessary information or to unmask specific fields temporarily under controlled conditions.
Data Source
AI summary
A co-browse service uses JavaScript to allow a web page shown in a user's browser to be viewed remotely. Updates to the web page are rendered into HTML and forwarded on the co-browse session. Aspects of the web page that should not be visible are specified in a list of masked elements which prevents the JavaScript from transmitting the content of those elements on the co-browse session. A person viewing the web page at the remote location can select objects to have those objects highlighted within the user's browser. Likewise the person viewing the web page may manipulate the objects by selecting objects and entering information into the objects. Updates to the web page are collected and aggregated such that only the most recent updates are forwarded on the co-browse session. Updates that don't affect the DOM, such as hover state, are also transmitted on the session.