Advances in deep learning enable models which can better cope with nuanced variations in text. In this talk, we explore how this ability to cope with fuzzy data can enhance cross-lingual name matching capabilities.