Lake Bled, one of Slovenia’s most visited natural landmarks, has experienced persistent algal blooms over the past century, despite multiple rehabilitation efforts. To support improved lake management, we conducted a national research project (V2-2370) focused on assessing the impact of pre-baiting fishing practices and other activities on lake water quality. An extensive database was compiled, combining historical data and new field measurements on nutrient loadings and the ecological and chemical status of the lake and its inflows. Within the project, we developed a catchment-scale GWLF model that simulates hydrological and nutrient-leaching processes to estimate long-term nutrient loadings. These outputs were used as inputs for a lake nutrient-cycling model implemented in AQUASIM, assuming phosphorus as the primary limiting factor. Additionally, an automated modeling approach using the ProBMoT tool was tested, which applies a declarative formalism for describing the system. It utilizes a domain-specific modeling library to induce models from data. At first, this approach was only applied to the lake modelling part, using the already established aquatic ecosystem modeling library, compatible with ProBMoT. In the second stage, the proposed approach will also be applied to the catchment modelling part using the pre-established watershed modeling library. Preliminary results indicate that the Mišca tributary is the dominant source of phosphorus and that internal lake loading significantly contributes to algal blooms. Future steps include exploring the advantages of automated modeling of Lake Bled within the AI4Sci Gravity project, in which we would like to connect both above-mentioned knowledge libraries. This integration will enable simultaneous automated modeling of nutrient loadings and lake water quality. The proposed AI-driven approach will support scenario testing and informed, model-based lake management strategies.