Abstract
The Czech Companion demonstrates that recent advances in Czech language: Automatic Speech Recognition (ASR), Natural Language Understanding (NLU), Natural Language Generation (NLG) and Text to Speech (TTS) synthesis can be combined with existing dialogue management (DM) techniques to build a dialogue system for natural-sounding dialogue. Certain components specifically developed in the Companions project surpassed the best existing results, as measured by the standard quantitative metrics used in the field (e.g., Czech and English full POS and morphological taggers, etc.).