The AI character is configurable across a range of behaviours and difficulty levels. At the most challenging setting, the AI deflects responsibility, minimises the incident, becomes confrontational, and responds to the manager's specific words — not a scripted sequence. The result is a conversation that feels genuinely unpredictable, mimicking the realistic pushback managers encounter when addressing difficult situations with resistant employees. The difficulty can be reduced for early-stage practice or increased as capability develops.

At the end of each conversation, the platform generates a detailed scored evaluation across multiple management competency dimensions — including clarity and directness, understanding the underlying cause, emphasis on safety and culture, accountability, action planning, communication skills, and policy awareness. Each dimension receives a rating and a written analysis specific to what the manager said in that session. This is not generic feedback — it is a precise account of where the approach was effective and where it fell short.

Yes. The admin dashboard allows organisations to create custom scenarios with their own settings, context, learning objectives, and AI character configurations. Evaluation criteria are also configurable per scenario, so the competencies being assessed can be aligned to your management framework, safety standards, or regulatory requirements. Scenario translation is built into the admin workflow for multilingual deployment.

Evaluation data from every AI Coaching session feeds into the FabricAcademy analytics layer, making management capability trends visible at team and organisational level alongside all other performance data. Coaching session outcomes can also inform Blended Learning pathway adjustments — if a manager consistently scores low on a specific competency, the platform can surface relevant learning content automatically.