Využití AI a Balanced Score Card v SMEs

This thesis proposes, tests, and evaluates the Augmented Strategic Mentor (ASM) model — a methodological framework for augmenting a consultant with generative artificial intelligence in order to make the Balanced Scorecard (BSC) accessible to small and medium-sized enterprises (SMEs). The research addresses two structural barriers to BSC adoption in the Czech context — the time intensity of traditional implementation and the absence of a structured facilitator — and fills a research gap in the previously uncharted intersection of BSC, generative AI, and the SME context. The method is a Single Case Study conducted in an anonymised printing company. The ASM cycle proceeded in five phases — from pre-diagnostics through the development of a strategy map and KPI matrix to a structured stress-test — while maintaining the Human-in-the-Loop (HiL) principle. The complete BSC cycle was delivered in 10.4 hours of the consultant's effective analytical work over 21 days, representing an order-of-magnitude difference relative to baseline estimates reported in the literature (RQ1). During the synthesis of a heterogeneous data source, the AI identified two hidden strategic capacities of the Slack Resources type and a factual inconsistency in the financial data; a stress-test conducted by three opposing models resulted in 13 accepted revisions to the BSC (RQ2). The thesis empirically demonstrates that the HiL principle is not merely a methodological postulate but a functionally necessary mechanism of output quality.