By providing various services, such as Incentive-based Demand Response (IBDR), buildings can play a crucial role in the energy market due to their significant energy consumption. However, effectively commissioning buildings for such desired functionalities requires significant expert knowledge and design effort, considering the variations in building dynamics and intended use. In this study, we introduce an adaptive Data-Driven Prediction (DDP) layer based on Willems' Fundamental Lemma to account for slowly time-varying building dynamics. This layer is integrated into a bi-level Data-enabled Predictive Control (DeePC) structure to achieve diverse control objectives. We validated the proposed method through a building-level case study involving participation in a Swiss IBDR program requiring early bidding, conducted in a real building testbed, Polydome. The adaptive DDP was utilized to develop a hierarchical controller that provides secondary frequency control on the demand side, with each layer designed to meet specific operational goals. Extensive testing with operational data from the Polydome demonstrated that the adaptive DDP improves prediction accuracy and reduces tuning effort compared to standard DeePC methods. A 52-day continuous experiment in the Polydome, using the tuned parameters, showed that the proposed controller achieved a 24.74% reduction in operating costs compared to a conventional control scheme. Our findings emphasize the potential of the proposed method to reduce the commissioning costs of advanced building control strategies and to facilitate the adoption of new techniques in building control.