Dilan,
It sounds to me like you are most of the way there.
From your description, you have 3 entities involved in this scenario: Account, Supervisor_Funds and Total_Funds. Allocated_Funds has 1 or more objects for each Supervisor and Total_Funds should have only one object for each supervisor. Total_Funds entity is populated via an after commit microflow on the Total_Funds entity.
If that is accurate, in your after commit microflow, you should first select all of the Total_Funds objects associated with the Supervisor for the Allocated_Funds object that has just been committed. Once you select all of these objects, you should delete them. In this way, your after commit microflow will always get rid of existing Total_Funds objects before creating a new one. Are you doing this in your microflow? If you are already doing this, perhaps you can post a model share of the after commit microflow so the community can have a look?
Also, bear in mind that you'll need to implement an after delete microflow if you haven't already so that your totals stay up to date in the event that you delete an Allocated_Funds object.
Hope that helps,
Mike
****UPDATE*****
Dilan,
You correction is accurate, my mistake. I have constructed a small test scenario that results in this page:
The data model looks like this:
Supervisor_Funds entity has 2 events: an After Commit and an After Delete. Each event calls the same microflow. That microflow looks like this:
You can access the microflow via this model share: https://modelshare.mendix.com/models/db47db9a-3107-4382-9a4f-25d645c90af6/after-commit-total-funds
Hope that helps,
Mike