The B2C Voicebot and the talking user manual – a practical example

In this article, I present a use case that really impressed me when I heard it. When Ralf Mühlenhöver from VIER GmbH told me about the case, I had to ask twice whether the solution really works as described. And yes, it is a reality and is used in practice by customers. Enjoy reading.

The challenges facing the solar energy industry

On the one hand, the solar energy sector is being driven by the trend towards renewable energies, but on the other hand it is facing a wide range of challenges: a lack of skilled personnel, aggressively priced competitors and long delivery times, as well as the unclear situation regarding the energy transition. All in all, a dynamic situation that has led the various departments of the company I am presenting here to consider how (generative) AI can help to better master the various requirements.

The company SolarEnergiePlus (original name has been changed) decided early on to pave the way for advanced technologies and to proceed iteratively. This means that the company begins to introduce new technologies in small steps in individual areas, then checks the success of each one and then expands the individual use cases step by step.

The company took the first step with the introduction of the “Sunny Solarius” voice bot based on ChatGPT. This innovative approach first revolutionized the company’s own customer service – and then significantly improved the efficiency and accuracy of solar system installations, among other things. The iterative approach specified by the company was thus rewarded.

Customer service redefined: Maxi Mustermann’s path to the optimal solar system

SolarEnergiePlus already had a classic IVR system (voice portal), which was able to put callers through to the right department based on their input. A quantum leap should be achieved with modern, generative AI. The experience of one of the company’s potential customers, let’s call her Maxi Mustermann as an example, illustrates the seamless and user-friendly process that the Sunny Solarius voice bot makes possible. Ms. Mustermann, looking for a solar system for her own home, used the telephone advice service offered by her local energy supplier and was immediately put through to the bot of the installation partner SolarEnergiePlus.

Sunny voice bot: “Hello, I’m Sunny Solarius, your voice bot from SolarEnergiePlus. How can I help you?”

Maxi Mustermann: “Hi, I’m Maxi Mustermann and I’m looking for a suitable solar system for my own home and don’t really know what size I should choose, whether I need a storage tank and what else I need to consider .”

Voice bot Sunny: “Hello Maxi. I’ll be happy to help you with that. Together we’ll find the right solution for you.”

Sunny Solarius asks Maxi various questions, such as roof area, house orientation, current electricity consumption, use of an electric car, current and estimated future electricity consumption, etc. Based on the information provided, the bot creates an offer in the background, which it sends to Maxi by e-mail. At the same time, the bot uses a workflow process to create a data record in the CRM system, which can be accessed by the experts at SolarEnergiePlus and the local energy supplier.

This dialog demonstrates the bot’s ability to respond individually to the customer’s needs by asking specific questions about the important information needed to prepare a quote. Based on this information, Sunny was able to create a customized offer in the ERP system, which was immediately sent to Ms. Mustermann by e-mail (robotic process automation and workflow orchestration are the technologies used here). At the same time, a corresponding data record was created in the CRM system, which ensures efficient further processing.

Maxi Mustermann’s positive surprise at the attractive offer leads to an order being placed by e-mail. Thanks to the integrated ACD (Automated Contact Distribution) technology, the e-mail is analyzed, immediately stored in the system and then presented as a record to the appropriate advisor, with the system prioritizing the process based on the recently created offer. This seamless integration of voice bot, email bot and CRM system illustrates how digital technologies can improve customer service in the solar industry.

Even more application possibilities: Support for fitters with Large Language Models

However, the benefits of artificial intelligence are not limited to advice and customer service. After the first positive experiences with the end customer bot, internal departments also contacted IT with questions as to whether their ideas could also be implemented.

One of these concerned the HR team, which was tasked with evaluating incoming applications and writing the “perfect” job advertisement – namely one that takes the different target groups into account when addressing them. Measurable successes were demonstrated by A/B tests.

Another department was the assembly team. This gave rise to the idea of how the technical staff could be better supported on the day of installation, as there is a shortage of experts in the back office who can provide live support, as well as fitters with many years of experience who can train new colleagues over longer periods of time.

Large Language Models (LLMs) now provide indispensable support for technicians on site. On the one hand, they combine global knowledge with specific information from the company, for example from knowledge databases, FAQs, operating instructions, assembly instructions, etc. Confronted with a challenging situation, the fitter, referred to here as Manni, used his service app to initiate a phone call to the service center. However, the call is taken by Knut, the installation advice bot, and no longer by one of his colleagues. The bot is also available to fitters for further support, serving as an interactive instruction manual and guiding them through specific installation steps. Questions about processes, spare parts, expansion options, etc. are also answered quickly and easily – and thanks to the deliberate use of the telephone interface, the staff on site have both hands free to carry out the work. This advanced solution ensures that installation teams can work independently with minimal external support, significantly increasing efficiency on site. The scenario underlines how the integration of AI into a wide range of processes offers an effective and efficient solution to technical challenges.

Conclusion

The introduction of the Sunny Solarius language bot and the use of Large Language Models to support the installation teams at SolarEnergiePlus represent significant progress not only for this company, but for industrial companies in general. The technologies have made it possible to personalize and improve customer service while making installation processes more efficient. Maxi Mustermann’s experience illustrates how the use of artificial intelligence not only increases customer satisfaction, but also improves operational efficiency, which will give SolarEnergiePlus a significant competitive advantage in the industry.

Did you like the article?

This use case is part of my Generative AI Trend Report 2024. In addition to many facts and trends about Generative AI in the DACH region, the report also contains many interesting applications of Generative Artificial Intelligence in companies.