As far as I know, groups and pools are job settings, and cannot be applied to individual tasks.
You can, however, submit a new job with just the task in question, and make it dependant on the primary job. From within your second/post job then, you can figure out which job you depended on, and grab whatever info you need from that job object if you need to figure out which files were rendered or something similar.
If you’re automating your submissions, this would be reasonably easy. If it’s a manual thing, a bit more messing around.
Common scenario for me is:
<Batch job group>
1) Processing job
2) Render job (depends on #1)
3) Slapcomp and notify (depends on #2)