A Turkish to Arabic translation practice project — what DeepL's machine translation got right, what I had to fix by hand, and how I handled a client glossary mismatch and a regional-fit flag along the way. Fictional client, real translation decisions.

I ran a fictional Q3 business report through Bing's machine translator, then sat down with it in memoQ to see what a full post-edit actually involves. Here's what raw MT got right, what it got wrong — a mistranslated fact, a tone shift, some terminology drift — and how I worked through fixing it.

A practical English-to-Turkish machine translation post-editing project covering MTPE, linguistic QA, cultural adaptation, and proofreading.