Localization – day to day issues, problems, doubts. Hoping for you input

Hello everyone,

My name is Jakub and I am a translator and localizer entering the world of ID and e-learning development.

As part of my post-graduate studies, I would like to create a short course, training material or similar tool that would help people from the e-learning world with localization.

The voices I've heard so far point at the fire (run DeepL API!?) and forget tactics many devs apply, hoping for the best and often putting their clients and learners in trouble.

I've recently re-read the well-known cliché that people need to Maslow before they Bloom.

What do you need – and don't have (enough of) to localize the carefully designed learning experience, not just the documents, audio and video?

For obvious reasons, MT/AI issues come to mind.

The problem with machine translation and AI-based translation is what we don't know. No matter how (relatively) good it looks at first glance, these things make random mistakes, hallucinate and operate on generalized assumptions.

I would really appreciate your examples and feedback on anything and everything that bothers you when e-learning and localization come together!