This project aims to build and test the prototype of a novel database management system for Knowledge Graphs. The novelty stems from using Machine Learning models trained to approximate expensive graph operators. Furthermore, we aim to give control to the end user over the uncertainty introduced in the query by the ML approximation via error bounding methods at the level of each query. We deploy and serve the ML models using ONNX, which was identified by our previous NAISS project as the most suitable tool for this case. In this project, we materialize the development of the vision presented in our published research papers: S. Horchidan, supervised by Paris Carbone. 2023. Query Optimization for Inference-Based Graph Databases. In Proceedings of the VLDB 2023 PhD Workshop, co-located with the 49th International Conference on Very Large Data Bases (VLDB 2023) [BEST PAPER AWARD] (https://ceur-ws.org/Vol-3452/paper9.pdf) S. Horchidan, and P. Carbone. 2023. ORB: Empowering Graph Queries through Inference. In Proceedings of 1st International Workshop on Data Management for Knowledge Graphs (DMKG ‘23), co-located with ESWC 2023. (https://ceur-ws.org/Vol-3443/ESWC_2023_DMKG_paper_6223.pdf)