MAGELLAN TV

For three years I translated documentaries from English into Arabic for MagellanTV, a niche documentary streaming service. The work spans five major genres: science, history, true crime, medical, and legal. Over 300 hours of finished content carry my translation.

Documentary translation is not interchangeable with general translation. It is technical, dense, and unforgiving. A science documentary on quantum mechanics demands precise rendering of terms that have settled English conventions but contested Arabic equivalents. A medical documentary requires anatomical and pharmacological vocabulary that the general translator does not carry. A legal documentary turns on jurisdictional language that does not map cleanly across systems. The work taught me to read for genre as much as for content.

I used SDL Trados Studio throughout. The translation memory and terminology bases I built across MagellanTV's catalogue are still in active use. The discipline of maintaining consistency across episodes, seasons, and adjacent series is what separates a translator from a localization engineer. The latter is closer to what the work actually required.

The pay was modest. The training was not. Three years of high-volume documentary translation across five technical fields is, in retrospect, a graduate-level applied linguistics program with a stipend. It is also the experiential ground from which I can now speak with confidence about what current AI translation systems get right, what they get wrong, and where the gaps sit. That informs how I think about The Architect AI's bilingual curriculum and the AI literacy work I do now.