AWS Transcribe, WebRTC

I'm working on project that involves AWS Transcribe. It's used to record interviews. We managed to get the transcription working but there's a couple of issues. 1. It's not transcribing in proper sentences - sometimes it doesn't use capital letters at the start of sentences and it's not adding question marks at the end of some questions, for example. Also, it's not auto recognising two different people when two people are speaking and so the transcribed document is confusing. This is a software product designed in React JS and Python and Django. If you know exactly what you're doing, I'd love to hear from you. This project isn't going to be for someone who can "figure it out as they go." We can do that ourselves but want to try and speed things up by hiring an expert.