Agent augmentation: Analyzing a customer call in real-time and suggesting the proper solution.


By loading the video, you agree to Vimeo's privacy policy.
Learn more

Load video

In case of bad video quality, manually change the resolution to 720p or higher on the video menu bar (settings icon)

Detecting an intent without learning from training data

In this insurance case an insurant explains to an agent on the phone that there is some problem with their car (the "A6"). The insurant continues to talk about their holiday in Portugal, before mentioning huge forest fires. Our AI understands the meaning of "forest fires" and that it has a negative effect on things like a car.

Integration into the existing ticketing system

Our solution is integrated into the existing IT infrastructure via REST-API. The agent is supported with the relevant client information on the screen and simply picks one of the suggested business cases that DEEPSEARCH provides.

Transcript of the analyzed call

Customer Service, my name is Horst Artwenger how can I help you hello my name is Huber I have a big problem with the car that I financed though your company can you please tell me the license plate number it is Salzburg Test 5 thank you Mrs. Huber I can see your data on my screen now what exactly happened I was in Portugal last week on holiday with my familiy and there were these big forest fires and we were right in the middle of it thank God nothing happened to us but the A6 really looks bad is the insurance going to pay for this

Proposed Solutions

DEEPSEARCH automatically detects the proper solutions that apply to the situation. It weights the cases with probabilities and ranks them from highest to lowest. In the video shown the two highest ranked solutions are properly identified:

1. Notification – Damage event

2. Notification – Write-off