Page 116 - Azerbaijan State University of Economics

Ilkin Seyidzade, Rudnak Ildiko: General Implementation Processes of Artificial
Intelligence and Its Economic Effects in Hungary

AI implementation steps
For companies to keep up with their competitors and gain a market advantage,
innovative technology must be integrated into the existing system. This can be a part
of a given activity task or sub-process, or it can be the automation of an entire
company and its activities. However, the implementation steps are worth going
through, whether it is only a small project or a larger project (Kessler and Gwozdz,
2018). According to Gassman et al (2017), developing an AI-based business model
is a series and set of interdependent activities aimed at creating enterprise value by
defining specific goals and improving performance.

Firstly, it is worth getting to know what AI and what type of AI is requested by the
company that wants to integrate the technology into the processes. What technology
is needed, what resources, and what the purpose of implementing AI is? Almost all
the information and online resources available today are at our disposal. It is
necessary to recognize and identify the problems that the enterprise is facing and to
identify the AI solutions that can be associated with the company's activities.
Financial and business values must be distinguished and these particular values must
be prioritized. In some cases, matrixes of potential and feasibility need to be ranked,
and the goals and priorities recognized by the management of the company should
be emphasized. Of course, there is a difference between what a company wants to
achieve with AI implementation and what it can achieve based on its capabilities and
resources. Identifying and organizing AI processes is the next step to eliminate
capability gaps and needs to be further developed internally. In case the company is
not familiar with the artificial technology used, it is worth considering to involve an
external expert (Marvin and Horowitz, 2018). Before the implementation process
begins, the definition of some milestones and metrics are needed. At this point, it has
to be determined at what points the company wants to improve and what measures
to improve. Defining indicators and forecasts will make the back-testing process
much easier later (Kessler and Gwozdz, 2018).

Many people may think that there is a standard business model for AI
implementation that is always applicable. In reality, however, much depends on
what activities and when the company wants AI-based technology. Depending on
the nature of the AI solution, it is based on huge amounts of data and algorithms.
Besides, the implementation of the AI business model can be divided into two parts:
 Infrastructure: providing IT services to others.
 Application: companies develop application services for clients based on
individual needs (Gonfalonieri, 2019).
3 AI solutions can be used, as shown in Table 1.

116