Learn more in the AI usage article.
How does Predictive Routing impact agent workload?
Predictive Routing can route more interactions to higher-scoring agents. Enable Workload Balancing to distribute volume more evenly. Learn more here.
Can customers enhance Predictive Routing with external data?
Yes. You can include external attributes as participant data to influence machine learning models in real time.
For predictive routing, with respect to AHT — will it look at AHT across queues for a given agent (CCR)?
No. AHT as considered by predictive routing is typically contextualized within each queue rather than aggregated across multiple queues for a given agent (CCR). This means that when predictive routing ranks agents for a queue, it uses the agent’s historical AHT where relevant for that queue, not a global AHT score from all queues served by the agent.
Does predictive routing take into account skill levels (i.e. prioritizing agents with full skills over those with “lower priority” skills) or do you need skill‑level reviews separately?
Yes. Administrators can turn skill matching on or off for individual queues that are configured to use the predictive routing method. Skill matching allows your organization to include agent skill settings in predictive routing decisions for inbound interactions. Skill matching includes all non-language ACD skills associated to the members of queues that use predictive routing.
Configure skill matching - Genesys Cloud Resource Center
Will we be able to run reports to see who is getting certain calls (i.e. to detect imbalances, training needs, etc.)?
Yes. The Queue Performance Detail view in Genesys Cloud shows which agents are handling interactions, with metrics like answered calls, handle time, and transfers. You can drill down to agent level and filter by time, media type, or skills.
For GPR-enabled queues, this view helps you see how predictive routing distributes calls across agents, detect imbalances, and spot where training or model adjustments may be needed.
🔗 Queue Performance Detail view – Genesys Cloud Resource Center
Could predictive routing potentially pull in quality evaluation scores(from managers) into the decision model?
No. Quality evaluation scores are not consumed by GPR models automatically.
Can you guard against the model favoring “too short” calls (i.e. does the model think 1–2 minute calls are optimal)?
Yes. The predictive routing model scores agents based on how likely they are to achieve the selected KPI, not just on call length. For AHT, the model learns from historical patterns — it recognizes that extremely short calls (like transfers or misroutes) don’t represent true success and does not prioritize them.
🔗 How the AI model scores agents for predictive routing
Please add your question to the post if not covered by the listed ones.