In this session we will review four years of an automated translation system as it matured from human translation, editing and proofreading (TEP) to neural machine translation (NMT) with post-editing (PE) and the changes in cost, throughput and technologies. We will also discuss lessons learned, challenges encountered and a current view of managing an on-demand platform.
Takeaways: Attendees will learn how to build a translation service architecture, how to convert a human translation process to a machine translation post-edited process and how to have fun with APIs, bots and developing automated systems.