Holiday Home AG automated customer service across multiple communication channels with a multi-agent AI solution. The AI answers recurring inquiries automatically and thus relieves the team. At the same time, the response times for customer inquiries have been significantly reduced while fully complying with the applicable data protection guidelines.
Holiday Home AG brokers vacation homes, apartments and other vacation accommodations throughout Europe. Thanks to an international team, the company, which was founded in 2003, serves its customers in over 15 languages.
2003
founded
15+
Languages in customer service
0.png){kind=icon}
7
Countries active
Holiday Home AG receives numerous customer inquiries every day regarding bookings, cancellations, payment status, or additional services such as pets. Communication takes place in several languages (German, French, Croatian, and Czech), which further increased the manual effort required. Frequently repeated requests led to long waiting times for customers and placed a strain on the service team. Holiday Home AG wanted to respond to these inquiries more quickly and reduce the workload on its employees. It therefore sought an AI-based automation solution that could be seamlessly integrated into its existing systems in compliance with the GDPR.
"CorpIn designed a customized multi-agent AI for us, which went live as a pilot within a few weeks. Today, it automates around 70% of our email inquiries and saves us valuable customer service resources every day."
Alexander Claashen
Managing Director Holiday Home AG
The project began with an analysis of the existing customer service processes and data, taking data protection requirements into account. This was followed by the definition of relevant use cases and the classification of query types as the basis for AI training. A modular system architecture was designed, including technologies such as LangChain, Rasa and OpenAI's LLMs. An AI prototype was then developed for the email channel, which processes frequent requests automatically and can trigger actions in connected systems via interfaces. The pilot phase included close monitoring and evaluation of the processing of customer inquiries to ensure response quality, as well as the implementation of a human-in-the-loop mechanism for complex cases. At the same time, the concept for integration into Holiday Home AG's existing CRM systems was developed.
The system already demonstrated strong performance in the test center: the average response time to email inquiries was just a few seconds, with a recognition accuracy of over 90% for predefined use cases such as "add pet," "cancellation," or "payment status." In over 80% of cases, the AI provided technically correct and fully usable responses that could have been sent without human intervention. For more complex or critical inquiries, a human-in-the-loop process is triggered, automatically involving an expert from the service team. Even under high load with up to 500 parallel inquiries, performance remained stable, with only slightly increased latency. At the same time, the running costs were extrapolated in detail from the AI test center to the rollout.
On this basis, a consistent, functional pilot was created – from the concept and technical architecture to the rollout plan. Today, the solution is in productive use: Frequent email inquiries are automatically processed by the AI agent, which significantly reduces response times for customers and relieves the service team. This leaves more room for quality support. The AI agent is also able to independently perform actions in the system, such as updating bookings on behalf of customers.
In the medium term, the concept of the AI solution is to be extended to other communication channels. The next steps on the roadmap are the integration of WhatsApp as an additional channel and the introduction of a voice bot. Thanks to the prepared CRM integration, the AI agent will be even more deeply integrated into internal processes in the future and will be able to handle customer concerns across all channels. Continuous monitoring and fine-tuning will remain important in the future to ensure reliable, transparent customer service.
