The increasing use of digital visual formats, especially educational videos, is making the transcription and automatic translation of videos one of the most demanded services in the sector. The time used for manual transcription is usually 3 or 4 times the duration of the video, adding also the creation of subtitles and the translation and revision of the content in several languages. Automatic transcription can speed up this task by automatically writing content and thus reducing work time. However, a poor automatic transcription may entail more work for those who have to review the transcription and, hence, translations errors. The TTT software developed by the UPV consists of advanced automatic speech recognition and translation systems. One of its advantages is its ability to adapt to the thematic area of each video, surpassing the solutions that generalist speech recognition systems can provide. In this way, the system recognises the context and automatically processes specific expressions of a specific field, which facilitates the transcription task. Thus, in the field of educational videos, for example, it offers a more complete service than that provided by YouTube, thanks to an extensive corpus that, just in Spanish, already includes more than sixty million phrases. The system can be used on-line on the portal https://politrans.upv.es/.