What's your current localization workflow?

We're seeing teams use everything from spreadsheets and ChatGPT to DeepL, Crowdin, Lokalise, and fully manual workflows.

What does your localization process actually look like today?

Always interested in seeing how different teams approach the problem.