From the Case to the Training Floor
The Agent node holds who the customer is and what went wrong, such as a parcel that never arrived, a double charge or a router that drops the connection every evening. A knowledge base adds the order history or the manual they are stuck on, which the trainee finds out by asking, as on a real call. With emotions, events and actions, this makes the following possible:
- •The customer's mood changes as the conversation goes, from Angry towards Reassured or further into Upset, and every line of the reply carries its emotion.
- •Event logging tells the customer what happened, such as the agent putting the call on hold or issuing a refund, and the customer can bring it up.
- •The action system lets the customer return an action such as ASK_FOR_MANAGER or HANG_UP that your simulation carries out.
A conversation flow scripts the call with the stages your process expects, and each checkpoint the trainee reaches comes back as a keyword your app can store.
Hard Calls Before Live Calls
New agents meet the angry, confused and impatient customer in training, not on day one of the queue, and can repeat the same case until they handle it well. Cases come from your own tickets and manuals, and group chat can add a simulated team lead for escalation practice in the same call. Checkpoint keywords and the conversation history from the Web API show where each trainee stopped. Neocortex does not grade the trainee, so how you review is up to you.






























