Automating Markdown and Image Translations with Co-op Translator

The Phi-3 Cookbook case study was the first public sign that my work was moving beyond individual AI application tutorials.

Before this, most of my writing focused on how to build with Azure AI: RAG systems, fine-tuning workflows, Prompt Flow, and environment setup. Those topics still mattered, but this project made a different problem visible. Technical knowledge does not only need to be generated or explained once. It has to be maintained across files, images, languages, and source changes.

The hard part was not simply asking a model to translate Markdown. A useful workflow had to preserve document structure, keep links meaningful, handle text inside images, organize output folders, and make the process repeatable inside a real open-source repository.

That changed the way I thought about documentation. I started to see translated docs less as content copies and more as maintained software assets. Once that clicked, translation became an engineering problem: state, structure, reviewability, and repeatability mattered as much as fluency.

This is why the post still feels important to me. It marks the point where Co-op Translator stopped being only a translation helper and started becoming infrastructure for technical knowledge.

Read the original article:

Automate Markdown and Image Translations Using Co-op Translator: Phi-3 Cookbook Case Study