Cancer research has been transformed in recent years by the considerable increase in omics data accumulated in public databases through the use of high-throughput technologies for profiling patient cohorts. The challenge now is to translate the molecular characteristics of a patient’s tumor into an appropriate therapeutic choice applicable in the clinic. Among the arsenal of anti-tumor treatments, targeted therapies and immunotherapies are now considered relevant therapeutic options used in first line for several types of cancer. However, a fairly large proportion of patients do not respond to these treatments or quickly develop resistance. The implementation of personalized medicine in our hospitals aims to help doctors better diagnose and treat their patients, by adapting the therapeutic choice to the molecular and cellular characteristics of each patient's tumor. Artificial intelligence (AI) algorithms are powerful tools on which to rely to advance precision medicine. However, these AI-based approaches are often “black boxes” regarding the reasons leading to a decision, penalizing their use in the clinic. The KATY project seeks to build a precision medicine platform based on AI systems. It will be hosted on a European computing infrastructure and accessible by several European hospitals. This platform will not only be efficient, but above all transparent when it comes to the molecular, cellular and clinical evidence underlying the recommendation of drug treatments adapted to each patient. Clinicians will be able to trust, evaluate and effectively use this AI system in their daily work. The platform implemented by the KATY consortium will be built around two main components: a distributed knowledge graph (DKG) and a collection of explainable artificial intelligence predictors (XAIPs). While the DKG is an intelligent repository that stores vast multi-omics patient information, as well as scientific information, the XAIPs will enrich the DKG and enable understandable personalized medicine decisions. This platform will be prototyped to predict the response of patients with kidney cancer to targeted therapies and immunotherapies.