This website aims to introduce my updated knowledge base and skill sets in Advance Translation Technology (ATT).

Over this spring, I initiated a pilot project training our own neural machine translation (NMT) with SYSTRAN Model Studio Lite. Our aim is to improve the French-English MT performance in translating annual reports of the Human Rights Council for the United Nations Office of the High Commissioner for Human Rights (OHCHR). 

With MT being my focus, I explored more in evaluating the quality of MTs available in the market by machine metrics (e.g. BLEU, MQM) through MATEO. Based on this, I advanced to Okapi Olifant, trying to improve the accuracy of Customized MT through corpus data cleaning, the pain point in MT training. 

Thanks to the extra deep miles in MT, as a localizer, I've developed a new lenses into Translation Management Systems (TMS). Except for SDL Trados, I experimented with XTM, Lokalise and others. This has equipped me expertise in evaluation, selection and adoption of TMS. One of my achievements in this is  a proposal to evaluate and select a TMS for Translation by Design.