Graph embedding methods have gained prominence due to their crucial role in effectively leveraging graph-structured data for various AI and ML applications. This study aims to address the challenges posed by the complexity of graph data in tasks such as graph classification. Despite substantial advancements in graph theory and network analysis, there is still a need to develop robust and scalable embedding techniques that can convert intricate graph structures into meaningful, lower-dimensional representations. The proposed solution, based on well-known and understood metrics describing graphs, not only matches existing solutions but even outperforms them in certain aspects. The simplicity of its implementation allows for easy modification of the range of considered metrics, which opens the door to further exploration of optimizations between the accuracy of graph embedding and computational and time complexity. In essence, research in graph embedding methods is indispensable for unlocking the full potential of graph-structured data across a lots of domains. As the volume and complexity of digital data continue to escalate, the development of robust and scalable embedding techniques becomes crucial. By bridging the gap between raw graph data and actionable knowledge, graph embedding methods pave the way for a new era of intelligent systems capable of understanding, reasoning, and learning from interconnected data at scale.