Personalised cardiac models have attracted increasing interest from the clinical community over the past decade, however clinical validation is often restricted to few cases. This is due to the computational complexity of the models and of their personalisation, but also to the difficulty of acquiring the extensive datasets required for such personalised models. In this talk, I will present exemples of clinical applications where we leveraged on data science in order to increase the applicability and robustness of such personalised cardiac models. Machine learning methods have been developed in order to benefit from both the physiological constraints of the model and the efficiency of the learning.