The aim of the amalfi project is to investigate the effectiveness and efficiency of graph learning algorithms for assessing huge, interconnected cryptoasset transaction graphs. The project attempts to solve some of the challenges associated with cryptoasset analytics, such as the reliance on limited ground-truth data and the increasing complexity caused by the new links between ledgers and layers in crypto-ecosystems. Building a comprehensive cryptoasset knowledge graph, conducting a systematic review of graph learning approaches in this context, and introducing a unique multi-layer conceptualization for cryptoasset transactions are expected outcomes. This project addresses the crucial need for powerful analytical tools in the face of increasingly complex and evolving digital finance.