Rob loves machine learning on sequences. He does projects on trajectories in sports, on electrograms at the cardiology, on speech, financial time series and on language. Rob currently writes his thesis on unsupervised learning for speech and works for NLP startup Frosha