Extract & Deduplicate Emails to CSV

I have a batch of Word documents—roughly 250 email addresses at the moment—where every address is written as normal, selectable text. Your task is simply to pull each address, strip out any duplicates so every entry is unique, and hand back a clean CSV containing one column of email addresses only.

I’d like this first set turned around quickly; if it goes smoothly, new documents will follow and we can treat it as an ongoing arrangement. Accuracy is vital, as is basic respect for the confidentiality of the source files.

Deliverable:
• A single CSV file with one column of unique, properly formatted email addresses.