Company
Nayya
Duration
Ongoing
Year
2026
Context
The existing platform at Nayya relied on traditional page-based navigation and linear workflows. While functional, the experience required employees to understand complex benefits terminology and know where to go before they could get help.
Our goal was to evolve the platform into an AI-native experience where conversations became the starting point for navigation, decision support, and personalized guidance.
challenge
Moving to an AI-native experience required more than introducing chat.
The existing product relied heavily on linear user journeys that assumed employees already knew where to begin and what they wanted to accomplish. In reality, users arrived with different goals, levels of benefits knowledge, and varying degrees of urgency.
The challenge was to redesign the experience so AI could proactively understand user intent, personalize navigation, and orchestrate multiple pathways across the platform rather than forcing everyone through the same flow.
My Role
I designed new adaptive user flows, developed prototypes, and conducted user testing to evaluate how effectively the approach supported different personas and needs. I synthesized the findings to identify areas for improvement and build confidence across the team in the new product direction.
goal
Tested design
User testing showed that AI-generated Advisor Insights helped employees better understand complex benefits terminology and the reasoning behind their recommendations.
Key findings:
Participants valued clear explanations of why plans were recommended.
Personalized insights made recommendations feel more relevant and trustworthy.
Breaking down insurance jargon into plain language improved confidence and comprehension.
Participants consistently relied on comparison views when evaluating multiple plan options.
Key findings:
Users appreciated having quick access to compare plans side by side.
Testing validated making the comparison view the default entry point whenever users had multiple plans to choose from.
