Our client, a large contact-centre organization, onboards up to 500 entry-level employees each year through a 10-day program split between role-play and agent practice. They partnered with Intelbyte to make that practice time count.
The challenge
With only half of the onboarding program available for hands-on work, practice capacity was capped at around 15,000 hours annually, and coaching feedback varied by facilitator and by scenario. Two trainees could finish the same program with very different preparation.
The organization needed a scalable way for trainees to practise more often, with consistent, brand-aligned scoring, and without adding training overhead.
The approach
Intelbyte built a voice-based training agent using Copilot Studio and Azure AI Foundry: an AI coach that role-plays as a customer across a range of scenarios, grounded in the organization’s own guidelines.
Every session runs with live transcription, a scoring rubric, and structured feedback, so each trainee rehearses real conversations and sees exactly where they stand against the brand’s standards.
Results
The simulator delivers twice the practice time at half the cost, shifting onboarding effort from role-play overhead to consistent scenario practice. Scoring and feedback are standardized against brand voice guidelines, removing facilitator variability.
Transparent reporting gives leaders readiness visibility they did not have before: trends can be tracked by cohort, by scenario, and by competency.