The SCATE project aims at improving translators' efficiency through improvements in translation technology, evaluation of computer-aided translation, terminology extraction from comparable corpora, speech recognition accuracy, and work flows and personalised user interfaces. It is funded by IWT-SBO, project nr. 130041. http://www.ccl.kuleuven.be/scate/ Envisaged Project Results We present the envisaged results of SCATE, now the project is mid-term, with two more years to go. We have surveyed and observed translators with respect to the following aspects: human-machine interaction in post-editing, human acquisition of domain knowledge and terminology, and workflow usage and interface personalization. We are researching different computer-aided translation (CAT) technologies, such as syntax-based fuzzy matching and concordancing, tools for speedier and more consistent collaborative translation, automated term extraction methods from comparable corpora, and integrated models and domain adaptation for speech as a post-editing method. Concerning MT Technology, we are working on syntax-based transduction, taxonomy-based confidence estimation metrics and speech translation. For these purposes, we have developed the following resources: a taxonomy of MT errors and manually annotated corpus of MT errors. Concerning the user interface, we are developing new approaches towards visualisation of translation features and towards flexible user interfaces. By the end of the project, we intend to integrate most of these aspects in a demonstration system that translates from English to Dutch. We are interested in feedback from language service providers and translators: what do you consider useful and interesting – how can we improve your translation environment? The Flemish Agency for Innovation through Science and Technology, Strategic Basic Research.