Recent advances in direct electron detector technology combined with effective strategies for image analysis have enabled the routine use of single particle cryo-Electron Microscopy (EM) to determine the structure of a variety of protein complexes at near-atomic resolution. The increased availability and access to cryo-EM resources within the structural biology community, has highlighted the need for robust and automated workflows for data analysis that can effectively and rapidly convert raw data into 3D structures. Although many components of the data processing pipeline can already run in an unsupervised manner, there are still several steps where user involvement is required in order to produce meaningful structures. The identification of these bottlenecks is the first step towards achieving the ambitious goal of fully automating the structure determination process by cryo-EM. In this talk, I’ll identify some of the roadblocks that stand in the way of full automation in single particle image analysis and discuss strategies for streamlining and establishing robust workflows for high-resolution structure determination by cryo-EM.