Prudential: Making Enterprise AI Useful for Employees
Industry
Insurance
Client
Prudential
Platforms
Mobile
Date
May 2018
CASE STUDY BRIEF
Capability
AI transformation, Enterprise UX, Stakeholder alignment
Leadership level
Lead
What I led
Designed an IBM Watson employee knowledge hub for Prudential APAC, giving staff one plain-language route into HR policies, company information, personal records, and support.
Why it matters
Defined a search-led employee service that separated general answers from protected personal information and made the next step clear.
DECISION RECORD
Problem and stakes
Employees had to piece together answers across HR policies, company news, personal records, and support channels. A conversational interface would only help if its access boundaries were understandable.
Role and scope
Led UX strategy and end-to-end product design for the Watson-powered employee knowledge hub.
Key decision and trade-off
Separated open company knowledge from protected employee data, used plain-language search as the entry point, and retained visible navigation and support paths for questions the assistant could not resolve.
Systems and artifacts
Service model, enterprise information architecture, natural-language search flows, privacy-aware access patterns, and the employee knowledge-hub experience.
Related content
The service problem
Employees were searching across separate HR, company, and personal-information systems. The product needed to reduce that fragmentation without making sensitive records feel casually accessible through a chatbot.
The key product boundary
I separated general company knowledge from protected employee information and designed the access states around that distinction. Plain-language search could start the journey, but the interface still showed where an answer came from and when the user was entering a protected flow.
What I designed
The work covered the service model, information architecture, query and result flows, protected access patterns, and routes into HR content, company news, personal records, and human support.
Scope of the evidence
The supported claim is the product and service design. I do not claim a lookup-time reduction or adoption figure because those metrics are not verified in the preserved source.
