Genesys Predictive Routing
I helped Genesys visualise and communicate the value of its AI-powered Predictive Routing feature. The feature showed customers how machine learning improved their key performance indicators (KPIs) without exposing the proprietary model behind it.
The goal was to make complex data transparent, understandable, and actionable, reinforcing trust in the AI’s ongoing value.
Organisation
Genesys
My role
Product Design
Timeline
6 months
My approach
Understanding the complexity
Reviewed documentation, analysed product requirements, and mapped existing workflows to clarify how the model functioned.
Deconstructed the “as-is” experience and synthesised past research to identify where customers struggled to understand the AI’s decisions.
Created flows and narratives to visualise the full GPR story – turning technical behaviour into meaningful business insight.
Collaborative design and alignment
Brought together Product Management, Engineering, and Data Science early to ensure feasibility and shared understanding.
Used rapid iteration and low-to-high-fidelity prototyping to get concepts in front of teams quickly.
Focused on progress over perfection by creating space for feedback and refinement rather than design-by-committee.
User testing and validation
Co-facilitated usability sessions to validate direction and answer key questions:
– Was the data clear and relevant?
– Did the visualisation match user expectations?
– Was the experience intuitive and seamless?
Key insights
Customers preferred clear feature names over technical IDs.
They wanted visibility into non-traditional KPIs such as Quality.
Visual export options would help them share insights with stakeholders.
Outcome
Delivered an analytics feature that made AI-driven results clear and consumable to customers.
Helped users see the sustained business impact of Predictive Routing post-launch.
Built trust in the AI through clear, human-centred data visualisation.
Informed the product roadmap through synthesised user insights and prioritised recommendations.
Established a repeatable process for testing and communicating AI-driven value across future products.
