Modeling of data structures has always been an important topic in the discussions of software engineering practice. Recently, the idea of conceptual modeling has lost importance in these discussions. The fact that research in this area has not been pushed further a lot for the last decade can be considered as an evidence. However, this concept has great potential. Especially the idea of creating a paradigm agnostic model depicting facts of the real world – the so called “Universe of Discourse” – instead of concrete data structures following a certain logical data model makes it so powerful and valuable. Hence, it deserves further research to find best practices to utilize conceptual modeling effectively. The problems that discouraged software engineers from making use of conceptual modeling is that the models are hard to understand. Creating them is time-consuming, other stakeholders do not know what to do with them and creating the final data structures requires an additional process step. After all, it is mostly perceived as too expensive in time and money without creating an appropriate value.In this article, the existing approaches are examined to find out their weaknesses and the reasons why they did not gain a broader acceptance. Therefore, the important requirements that a conceptual modeling language has to meet for practical fielding are determined. Furthermore, the concepts of semantic modeling languages are examined. Using semantics instead of mere structural discussions simplifies access and understanding for non-IT stakeholders. It helps to check the validity of the created data structures against the demands of the real business. In the further course, the concept of semantically irreducible sentence modeling will be discussed which can act as a bridge between semantic and conceptual modeling.With the results of these discussions, the conceptual modeling language AGILA MOD is presented. This modeling language bases on the idea of depicting semantically irreducible sentences as graphical model. By this, it can act as a common platform all project participants can agree upon building the bridge between IT implementation and business requirements. The models can be created from semantically irreducible sentences and they can be read backwards into semantically irreducible sentences making this language easy to understand for all project participants. AGILA MOD is therefore intended to be as easy as possible to get started without a lot of learning efforts. Hence, it bases on the well-known Entity-Relationship language in a simplified variant. A few additional constructs are added that also refer to well-known modeling techniques reducing the efforts of learning new elements nearly to zero.The derivation of AGILA MOD models to a logical model is done by following simple derivation rules making it less time consuming and hence less cost-intensive. This language shall act as a basis for further research targeting towards the new logical models of NoSQL as well as creating a comprehensive framework automating the derivation as much as possible. Additionally, the possibility of making use of polyglot persistence with this approach and the creation of a convenient API shall be considered in future research.