AI Email Management

Improved AI Training for Email Labelling

January 27, 2026 Comments Off on Improved AI Training for Email Labelling

We’ve just rolled out an improvement to AI training for email labelling in ActorDo.

If you train Actor using examples, you’ll now see a new command line available:

Then mark as:

  • read
  • unread
  • important
  • not important
  • starred
  • unstarred

Why this matters

Until now, training mostly focused on how emails are labelled.
With this update, training also covers how emails should be treated after they’re processed.

That means:

  • Better prioritization
  • Cleaner inbox states
  • Smarter automation over time

You’re no longer just teaching what an email is —
you’re teaching what should happen to it.

Platform support

As always, AI training works across:

  • Google (Gmail)
  • Microsoft (Outlook)

No change in setup. Just more control.

This is part of our ongoing effort to make AI training feel natural, explicit, and actually useful — closer to how you’d explain things to a human assistant.

More improvements coming soon. Don’t forget, we build ActorDo as the best Work Assistant.

Hi, I’m ACTOR, designed to simplify life for busy professionals like you.

My mission is to be the best digital partner you’ve ever had: for keeping your schedule on track, managing your emails, or helping you focus on what truly matters.

Videos: getting started

I’m always learning and improving because I want to understand your needs better with each interaction.

Think of me as your reliable co-pilot, here to reduce your stress and give you back more of your precious time.
Together, we’ll conquer your to-do list, stay organized, and make every day more productive.

Let me handle the small stuff so you can focus on the big picture. Ready to make your day smoother? Let’s get started!

Try me out here

Note: The comparisons or information present in this article/page are based on publicly available information and may not be fully comprehensive. All product and brand names mentioned are trademarks or registered trademarks of their respective owners.