23 August 2019

Systematically Using Artificial Intelligence to Exploit Opportunities

DATEV AI (artificial intelligence) lab analyzes possible scenarios.

With the founding of an AI lab for artificial intelligence DATEV eG has set the course for an increasing use of AI in the company‘s software solutions. The newly created team of experts is positioned in the DATEV lab, acting internally as single point of contact. It coaches product managers from all special departments when it comes to AI, analyzes specific scenarios, and implements them prototypically.

AI will soon lighten our workload of routine jobs – even in office routines. Therefore, relevant technologies have strategic meaning for DATEV, too. First projects have been carried out already. For example, a solution automating parts of financial bookkeeping using AI is already running in a controlled test mode. In DATEV‘s service area AI helps handling user questions faster. Within those projects for self-learning systems it became obvious that building up know-how and skills in the AI context is a challenge.

Realizing and Implementing AI Potentials

Many applications for different purposes suited for using AI technologies may potentially simplify the use of DATEV Software: language assistants, chatbots, and knowledge management can help structuring data and making it available.

Using recognition and allocation of voucher information or semantic interpretation of invoices can raise the degree of automation in software. Its efficiency and quality can be improved with internal process monitoring and anomaly detection in the accounting process.

In order to realize such applicability all special departments can now turn to the AI lab. It analyzes to which extent AI technology is suitable for improving a product and coaches during its implementation. It supports product managers to identify AI potentials, evaluate application scenarios, and develop prototypes. Where necessary, the AI lab cooperates with research institutes or start-ups, it already has close contacts through the network TechFounders. Topic example for such a cooperation is an evaluation project for anonymization and pseudonymization of personal data. This feasibility study, realized with the start-up KIProtect, will provide a sound basis for data protection compliant analysis of such information – be it for development purposes or use in data-driven big data analyses.