Processing applications faster in the administration

Sifting through and evaluating files takes up a lot of time in public authorities. The complete processing of individual cases often takes a correspondingly long time. The example of the Berufsgenossenschaft Energie Textil Elektro Medienerzeugnisse (BG ETEM), a statutory accident insurance institution, shows how an AI-based system can speed up administrative work processes. 

A stack of bundled paper on a table can be seen in the foreground.
BG ETEM has been successfully using Artificial Intelligence within the company since 2019 to make administrative tasks faster and more effective. © smolaw/Shutterstock

Every year, around 180,000 reports on occupational accidents land on the desks of the employees in BG ETEM's accident processing department. Until recently, a large proportion of these, in particular the 55,000 or so reportable occupational accidents, were still checked by hand. The numbers give us an idea: A processing backlog is looming. One reason for this is the lack of personnel in public and private administration. A far greater one is the ever-growing flood of digital data. What if the work process could be accelerated with the help of learning systems?

AI optimizes the subrogation reporting process

BG ETEM implemented this wishful thinking. Since 2019, Artificial Intelligence has been making daily work easier for employees. The intelligent software automatically scans each new accident record and then compares the associated data with millions of occupational accidents that have already been completed. Within a very short time, the AI can provide a quick and straightforward recourse recommendation for each accident. Recoverable accidents can be quickly and effectively discovered and processed by employees.

Specifically, the self-learning AI software recognizes patterns in successful subrogation cases, applies them to new and unknown cases, and thus generates case-accurate forecasts of subrogation probabilities and accident costs incurred. The AI permanently produces decision tree models. Cost forecasts are thus easily comprehensible for all parties involved.

Decisions are still made by humans

In addition to rapid analysis, the AI system brings another advantage: Many a recourse case was not discovered during manual processing, resulting in the loss of a considerable amount of revenue. BG ETEM has already been able to save around 1.1 million euros by using the technology in the system's introductory phase since 2019.

Will the AI system eliminate jobs in the future? Jörg Botti, member of the BG ETEM management board, emphasizes that this is not the case - and that AI supports people in administration. "The decision for which cases to start a recourse procedure still lies with a human being, but is made enormously easier by the AI tool." In September 2020, BG ETEM was awarded first prize in the category "Best project for the use of innovative technologies 2020" for its system in the eGovernment competition.

Application facts

Technology field
Data Management and Analysis
Application industry
Finance, Insurance and Real Estate
Information and Communication
Field of application
  • Predictive Analytics
    Data Analytics
Value-added activity
Service/Customer Service
Human Resources HR
AI Developers

Berufsgenossenschaft Energie Textil Elektro Medienerzeugnisse