Yes please. It would be nice for celery to have the option to store the results as native json and not binary into mongodb. There was a similar issue raised in the past but had no pull request #1200.
This pull request seems to address it.

This address the first request of #1200, the second one is not really needed, since you would need the object id nevertheless (or task id for results), so the best option is to save to a collection all the data we need before the task is started and reference this in the results with its object id.

Thanks! Merged into master.

I don't think you need to register a kombu serializer for this, so I changed it to override the backend methods instead.

I also don't think one should expect to get stores group results by doing AsyncResult(group_id).get().

Can you also add your name to https://github.com/celery/celery/edit/master/CONTRIBUTORS.txt?

Is there any reason this hasn't been backported to the 3.1 release? The part of the code that does

if self.serializer in BINARY_CODECS:
    payload = Binary(payload)

fixes this issue: #2081

actually, this still doesn't quite fix it

'result': self.encode(result) needs to be changed to 'result': result as shown in #2081 (comment)

@ask what are your thoughts on adding a configuration option allowing to store to content as a raw dict instead of doing self.encode(result)?

@AlJohri @thedrow Any update on this, or a new issue where we can discuss it? I definitely think having the raw json in mongodb makes sense, otherwise there isn't a whole lot of benefit to using mongo in the first place.