Very poor video/audio needs audio enhanced

We have a very poor quality cctv video file mp4 with big silent sections or background noise and some poor quality speech. We do not expect miracles but we need to remove the empty space and try to get the speech enhanced to identify any words possible.

It is less than 12 minutes with a small amount of speech. A transcript of identifiable words would be great. 48 hours maximum turnaround.

If we can identify the sound source such as tv sound etc or actual live speech, that would be a help.