The knowledge of the magnetic field structure of coronal mass ejections (CMEs) is crucial for understanding their early dynamics and eruptions as well as predicting their space weather effects at Earth.  For example, this information is crucial  for constraining the flux rope parameters both in first-principle space weather simulations (e.g, ENLIL, EUHFORIA and SUSANOO-CME) and in semi-empirical CME models. At the University of Helsinki Space Physics team we are developing both observational schemes and fully data-driven simulations to estimate the initial magnetic field in CME flux ropes. Observational scheme combines several indirect proxies, while our data-driven simulations uses a  time-dependent magnetofrictional modelling (TMFM) approach to estimate the field  self-consistently.  In this seminar I will discuss the “Bz problem” from a space weather point of view and present the above mentioned activities within the team.