Skip to main content

Projects descriptions for better AI tagging

Features & Improvements Request

The auto-tagging of my tasks is not working super well, having a way to define rules / some description per project would be a great way to help the LLM identify which project to tag. E.g. project “engineering improvements” I could put some examples (“find better way to do XY”,”prepare OKR process” or similar) or hints on when to apply this.

Posted by Alex ·
5 comments

Log in to comment and vote

Comments5

  • Abby Lupi

    •

    May 14

    •

    Merged request

    •

    1 vote

    Improved Auto-Project Labeling

    Ability to add 1–3 sentences of context/description to each project to help the AI better understand and correctly assign tasks.

  • Ryan

    •

    Feb 13

    I would also love project descriptions. Agree that they would be useful for AI, but I’m thinking them more as a place to capture notes, resources, and links for that project/goal/objective that are not tasks.

    • Akiflow Team

      •

      Feb 16

      Thanks for sharing your thoughts! Adding project descriptions as a space to capture notes, resources, and links, beyond just tasks is a great idea. We’ve taken note of this feature request for future consideration.

  • Tadej J.

    Team•

    Feb 3

    Hey Alex!

    Thanks for sharing this feedback. I'm Tadej, Product Manager at Akiflow, and we really appreciate you taking the time to explain how AI Project Assignment could work better for you.


    To help us understand the issue better and explore the right solution, I have a few quick questions:

    1. Are you manually correcting the mis-tagged tasks when the AI gets them wrong? If so, roughly how many per week are you having to fix?

    2. Have you noticed the AI improving over time for similar tasks, or does it keep making the same mistakes even after you've corrected them?

    3. Can you share 2-3 specific examples of tasks that consistently get mis-tagged? ("Task title" → AI assigned it to Project X, but should be Project Y)


    Understanding these will help us figure out whether we need to improve how the AI learns from your corrections, or if adding project descriptions/rules like you suggested is the right path forward.

    Looking forward to hearing more! 

  • Akiflow Team

    •

    Jan 29

    @Alex Prams I’m definitely passing this over to our product team. It’s a great shout—anything that cuts down on manual cleanup and makes auto-tagging actually 'auto' is a win in our book.

    In the meantime, the best (though admittedly annoying) workaround is to pack the task title with specific keywords to nudge the AI in the right direction. I know that’s not the dream setup, but it should help for now!