If you are getting a lot of emails into your Gmail inbox, then probably some of those can be solved easly by replying or by doing specific short tasks. But analyzing whole email thread content just to catch up with multiple threads can be very wasteful. So by using AI you can actually get simple propositions of what should be done before closing this specific email and actual proposed answer to that email.
This is especially useful if you need to do some actions before replying to email. In that case you can simply assign task to specific person, await until it's done, copy-paste AI answer when it's done, and close.
Another good use would be if on one inbox there are working multiple people. It can make the process much more streamlined.
NOTE Chat does not heave memory attached on purpose. The purpose is that it should analyze each inbox message separately, not in thread. When using memory, it can get lost easily.
NOTE2 You might want to adjust limits on nodes "Get Unread From Inbox", "Get Starred From Inbox" and "Get Open Tasks", especially if having issues with model complying to output structure.
And that's it. I hope that this automation will make your Gmail <-> Todoist process much more streamlined!
There is actually more that you could do with this automation, but it really depends on your needs. For example, you could add Form trigger to handle incoming support requests. Another thing is that you could replace Todoist with Asana or any database (like NocoDB) if you are using it for your task management.