 The paper introduces deep lacy, a software suite that uses deep neural networks to rapidly analyze single molecule data from experiments such as SMF-ET, automatically sorting recorded traces, determining fret correction factors, and classifying state transitions in 20 to 100 milisicons per trajectory. The technique was benchmarked using ground truth simulations and experimental data analyzed manually by an expert user, and compared with a conventional hidden Markov model analysis. Deplaces versatility is demonstrated through its ability to analyze both total internal reflection fluorescence microscopy and confocal SMF-ET data as illustrated using a highly tunable L-shaped DNA origami structure. This article was offered by Simon Wininger, Puglia Asadia Tuei, Johann Bolen, and others.