Die „Gov Pod“-Serie: Microsoft in der Regierung
Erfahren Sie, wie Cloud Computing, neue Technologien und Blockchain das Leben von Mitarbeitern und Bürgern verbessern können.
Jonas Huonder
Marketing Lead CMO, Microsoft Switzerland
As emails now make up a large part of communication, companies face the challenge of processing large volumes of messages efficiently. This is especially crucial in a 24-hour contact center like Mobi24, a subsidiary of Mobiliar, where quick and error-free handling of customer inquiries is essential. With the introduction of an AI-based email triage system in the summer of 2022, Mobi24, in collaboration with Microsoft, has taken a significant step towards automation and increased efficiency.
The decision to introduce an AI-based email triage system arose from the need to manage the ever-growing flood of emails. Isabelle Walther, a business analyst at Mobi24, explains: „With the increase in emails compared to phone calls, the pressure on our team to process these inquiries quickly and accurately increased. The introduction of AI triage was therefore a logical step to optimize our workflows“.
The solution was designed not only to handle the increasing workload but also to ensure that emails are efficiently forwarded to the right places for quick processing. „Our priority was a powerful and user-friendly solution for all Mobi24 contact centers,“ Walther adds.
The development of the triage system was carried out using Microsoft Azure. „Our AI models were trained on Azure Machine Learning and deployed via Azure Kubernetes Service,“ explains David Hess, a data scientist at Mobi24. „The Power Platform can query these services, enabling seamless integration into our existing workflows.“ Hess emphasizes an important point: „All data remains within the company, and no information is shared externally,“ ensuring the security and confidentiality of customer data at all times.
The multi-stage workflow of the email triage system first checks whether an email received via Outlook is a complaint. These emails are directly forwarded to the complaints management team. In the next step, the Power Platform identifies damage reports and forwards them to the responsible general agency. Other topics, such as contributions to the customer magazine Mobirama or medical reports, are also specifically recognized and forwarded to the appropriate departments.
„By combining AI-based analysis and rule-based categorization, we have now achieved an automation rate of 45 percent,“ reports Walther. „Our goal is to increase this value to 50 percent.“ Automation not only reduces processing time but also relieves employees of repetitive tasks, allowing them to focus on more complex and customer-oriented activities.
The development of the email triage system was a collaborative project from the beginning, involving individuals without extensive technical knowledge. A hackathon, where different software solutions were developed collaboratively in a short time, played a key role in this process. Isabelle Walther recalls: „I was involved from the beginning and worked on the configuration myself. It was a valuable experience and shows how accessible the Power Platform is even for non-technicians“.
The introduction of the triage system took place in several phases. During a test phase, triaged emails were initially moved to subfolders instead of being forwarded directly. This allowed employees to familiarize themselves with the system and provide valuable feedback. „Our goal was to build employees‘ trust in the system and ensure that the triage works smoothly and error-free,“ explains Walther.
Thanks to this cautious approach, there were no reservations among employees about the new technology. „Our team is very open to technology and welcomed the automation of a rather unpopular, mechanical task,“ reports Walther. The system works so unobtrusively in the background that it is often no longer noticed. „This illustrates how well the technology is integrated into our work processes,“ she adds.
The introduction of AI-based email triage at Mobi24 is an example of how technological innovations can sustainably improve the efficiency and quality of customer service. Through close collaboration with Microsoft and the targeted use of Azure technologies, Mobi24 has succeeded in developing a solution that not only reduces the workload of employees but also increases customer satisfaction. In a time when fast and precise communication is crucial, Mobi24 sets new standards in the insurance industry with this technology and remains true to its motto: „Proud to help!“.
The introduction of AI triage was a significant strategic step to optimize our processes and sustainably reduce the workload of our team.