can i add some sort of filter that it only recognizes my voice? i.e a pre-filter or smth?

what are you trying to do? check out the speechmatics plugin. it has real time diarization. you can instruct your LLM to pay attention to just one of the speakers. here's a nice demo of it and link to the plugin. https://docs.livekit.io/agents/models/stt/plugins/speechmatics/ https://x.com/Speechmatics/status/1963192047151812953

Making it only focus on my voice even in loud places not picking up other people in the same room

Have you tried the Krisp background voice cancellation? It can do a good job of filtering out background, but the processed audio sometimes causes more errors in the STT than just sending in the unprocessed audio to the STT. but it's worth trying for your use case. https://docs.livekit.io/home/cloud/noise-cancellation/#overview