AI, Blockchain, Chatbots - data quality first

The many current project examples of the insurance industry show that it makes no sense to develop the future if to the present day flaws exist. The industry currently works on glamorous topics such as block chain, chat bots and artificial intelligence. The point is that automated processes can only be as good as the data available to them. However, there is still a lot of catching up to do there.

The right hand needs to know what the left hand does

Contrary to the way they appear from the outside, insurance companies are often split into several companies. But even where this is not the case, IT systems are often located in different silos. Therefore, no comprehensive access to the stored data is possible. As a result, even if accident insurers and car insurers belong to the same group, artificial intelligence can only access a single collection of data. In this case data is intelligent, but not very educated. Legal framework conditions make it difficult to improve – and this is precisely where innovative ideas that are appropriate to the regulatory requirements are needed.

Block chain or what intelligent data management could enable

The term block chain stands for computer networks that can exchange data in a decentralized and supposedly very secure way. Here, not only comprehensive access is possible, but also fast changes to data records, no matter where they are stored. On this basis, so-called “Smart Contracts” can be generated – self-executing contracts that make it easier to negotiate, execute and verify contracts. In addition, data records from third-party providers can be integrated, such as flight systems. These would then report delays of aircraft, for example, so that an automatic damage payment can be initiated.

Walking before running

There is no point in dealing with hype topics until the necessary basis has been created. No one values “stupid” chatbots, which only demonstrate that the idea is good, but the implementation is questionable. Insurers must first get their product digitization and data storage under control.

This article was first published on LinkedIn.

Share