Improve phone support by 355% with AI

AI based telephone support for happier customers and better service

Summary

• Telephone support 24x7 with interactive conversation management

• Handle 355% more requests with the same workforce

• Integrate real-time data to answer questions

Challenge

The company "Entsorgung Meier" operates a call center with 5 employees, who primarily provide information about the company's services. This includes, for example, the opening hours and addresses of drop-off locations, permissible valuable and hazardous materials, and additional services such as home collection.

5 employees are continuously engaged in handling inquiries via phone from 8 AM to 4 PM. On average, each employee processes between 10-15 inquiries per hour. During peak times, there are regularly waiting times of 10-15 minutes.

Due to the ongoing challenge of attracting employees for the call center and offering this service beyond usual business hours, the availability of telephone accessibility has been progressively reduced. Since this directly affects the perceived quality of the service provider, a project was initiated to improve quality and extend accessibility through automation and intelligent use of technology.

Action

Based on the bloofactory Prozess , a prototype for an OpenAI AI-powered telephone hotline was developed within a few weeks. Calls are made via landline and enter an AI-supported holding loop. The AI interacts with the caller through voice and answers questions about opening hours, acceptance addresses, permissible valuable material products, etc.

The AI-based bot uses an existing and prepared knowledge base, which is also used on the website, to answer callers' questions accordingly. A key goal is to ensure a natural interaction process so that customers experience a good user experience.

Results

During a 3-month trial period, a sophisticated system was developed, and callers can initiate an interaction with the voice AI-based solution via landline call.

Customers can use natural language to ask questions. Dialogues in the prototype are possible in both German and English. The AI automatically recognizes the language being used and responds accordingly.

Within the test period, over 5,000 calls were handled by the AI. It was found that there was a potential cost saving of 88%. In other words, the existing employees were able to increase their effectiveness by a factor of 3.55, or 355%. At the same time, the service was expanded to 24 hours a day, 7 days a week, without changing the working hours of the affected employees.

Perspective

The AI's data foundation is continuously being expanded, and natural events such as absences due to illness, holidays, and other downtime are automatically taken into account. The voice telephony is further developed and additionally linked with a text chat. In the future, customers will be able to make calls and also book extra services such as container provisioning or arrange for pickup.

Furthermore, in the future, calls via WhatsApp and similar platforms will also be integrated.