Automating 90% of incoming requests
identification with a natural language IVR

Chief Technical Officer (CTO)

@ French Top Tier Delivery & Logistics Group

calldesk has been working since 2018 with the CTO of one of the largest delivery & logistics French players. This CTO wanted to facilitate its internal customers’ support concerning their requests for office and computer support, and to promote their autonomy.

Long waiting time due to numerous routing errors

Because of the company’s size and the many application services that are run internally, hundreds of thousands of employee’s phone calls are received each year by the service center.


Each group of agents has its specific field of expertise, and it is difficult for agents to handle all the increasingly diverse and complex requests. The service center does have a keypad DTMF IVR, but there are many routing errors due to callers finding the system heavy, complex, and therefore bypassing it.


As a result, agents who pick up calls are often unable to resolve incoming requests themselves, which multiplies internal transfers, leading to higher waiting times and satisfaction level falling.

The goal: to relieve the contact center and avoid routing errors

In order to be able to provide each request with an adequate response, the CTO is looking for a way to relieve his agents of part of the incoming call flow, while avoiding routing errors.

If it is possible to immediately identify each call’s reason more efficiently than with the DTMF IVR, simple and low value-added requests could be automated. Other requests coud more easily be transported to the appropriate skills center. The CTO has therefore used calldesk technology to achieve this result.

The result: 90% of requests identified and 50% selfcare rate

From the first months, the CTO was able to validate the solution’s ROI. Implementating a natural language IVR has greatly simplified the caller’s customer journey, who are directly directed to the right pole of expertise.


In addition, simple and recurring requests have been automated, reducing by half the number of calls handled by human agents in the service center. Month after month, CTO teams continue to identify new use cases that can be automated, and the voice agent supports an ever-increasing call flow.


Understanding rate

60 s.

Voicebot AHT


Selfcare rate


Routing errors eliminated

Discover calldesk today.

Talk to a customer service expert.