Answering every call would be a poor achievement if customers still could not get their problem resolved.
FLAI was the voice AI example we examined in May. Its automotive focus brought a practical question into the Omnia discussion: could a conversational layer help a dealership respond when employees are busy, while preserving the quality of the next action?
Vertex would evaluate the handover before celebrating the conversation. Has the caller’s identity been checked appropriately? Is the request understood? Has a booking been confirmed against real capacity, or merely recorded as a preference? A fluent voice should not turn an uncertain position into a confident promise.
The difficult cases define the operating model. A breakdown, a disputed invoice or an anxious customer needs an appropriate route to a person. The receiving employee should see an accurate account of the request and any commitment already made.
Voice capability remains a development proposition for Omnia. The FLAI discussion does not establish a supplier agreement, a production integration or evidence that an AI advisor can handle every category of call.
Before choosing a voice system, test five awkward calls alongside five ordinary bookings. Measure whether the next employee can complete the customer’s request without starting again. That is a stronger acceptance test than conversational polish.