Nürnberger Versicherung is an insurance company founded in 1884 and headquartered in Nuremberg. With almost 4,500 employees and approximately 6 million active insurance policies under management, Nürnberger generates annual sales of approximately € 4.5 billion, making it one of the most important insurers in the German-speaking region.
The project: Search by relevance – automatically analyze AVB inventory and extract migration-relevant modules
In migration projects, the general terms and conditions of insurance (GCI) have to be searched for migration-relevant information. This is a tedious and time-consuming step. The goal was to develop a prototypical tool that can automatically analyze the inventory of GCIs and extract or highlight migration-relevant modules.
The challenge
When migrating insurance portfolios, it is necessary for the insurer to be able to map the old portfolios completely in the new system. In order to obtain an overview of the portfolio, the general terms and conditions of insurance (GCI) must be searched for migration-relevant information. This requires a great deal of manual work on the part of the insurer.
The solution approach of Convista
A characteristic feature of AVB analyses is the large number of documents and formulations. In close cooperation with the experts at Nürnberger Versicherung, we manually analyzed a small sample of standard terms and conditions (approx. 40 documents) and marked the relevant components as such. This data was then pre-processed and made “machine learning ready” using a specially developed pipeline. On this data basis, various machine learning algorithms were trained and finally the best model was selected.
The benefit
With the help of the “Highlite” tool, the customer can analyze an arbitrarily large inventory of AVBs automatically and on-demand and extract the elements that are actually relevant for migration. The previously manual work step is automated and delivers a precise result. This saves time and effort. In addition, the output provides the opportunity to examine one’s own inventory of AVBs in detail, for example to analyze the differences and similarities between the various AVB generations.
Why was Convista the right partner for this project?
contact: Do you have any questions?
Feel free to contact us:
Your contact for the insurance industry
Matthias Müller