Návrh implementační strategie AI voicebotů v telekomunikačním odvětví

The aim of this bachelor's thesis is to design a quantified framework for assessing the economic rationality of deploying an AI voicebot in the Czech telecommunications sector and to determine the conditions under which such an investment pays off. The theoretical part integrates the framework of Baumol's cost disease, the architectural limits of large language models, human-computer interaction, anthropomorphism risks, and the regulatory constraints of the AI Act and GDPR. The practical part builds on a prioritization matrix of five use cases and develops a financial model combining the Forrester Total Economic Impact methodology with the Czech wage baseline. The model quantifies net present value across three scenarios at a 10% discount rate over a five-year horizon: conservative +USD 9.58M, realistic +USD 11.78M, and optimistic −USD 0.95M. The sensitivity analysis shows that conversational response latency is the dominant value driver. An integrated operational architecture and a Bass diffusion model simulate a sectoral inflection point between 2028 and 2029. The thesis formulates a four-phase implementation framework that serves as an analytical tool supporting the decision-making of Czech mobile network operators. The quantitative outputs of the model are not a prediction of the economic impact for a specific operator, but a scenario-based simulation of value range for a typical sector representative calibrated against the publicly available wage and regulatory baseline; primary empirical validation is recommended as a follow-up step.