Microsoft utility to convert PDFs/Word documents to Markdown: markitdown

I noticed that Microsoft has published a rather new (first commit in November 2024) utility on GitHub called "markitdown". It appears to be an useful little application: if can take PDF files, Word (.docx) and Excel (.xlsx) documents, and convert them to Markdown.

While this can be useful in itself (even Notepad can display them in a visual view instead of just plain text), I find one very good use for this: converting existing design documents (often, in Word format) into Markdown for AI agents to consume. I've found that especially PDFs can be tricky to AI, so a little help here goes a long way (read: saves tokens and thus, time).

The utility is written in Python, and GitHub has the source.