[LAD] Headless mix line in + Bluetooth

Hi everyone, my grandmother needs headphones to watch TV, but wearing them causes her to miss all the conversation around her. Hearing aids that allow both TV and microphone input are around 2.5k€. As I'm quite firm with Linux, and have conducted several Raspberry Pi projects, I wondered if it was possible to find a DIY solution for this. I thought of using regular Bluetooth headphones with an integrated microphone, and a Raspberry Pi that gets fed the TV audio. Now it should be possible to receive both the microphone audio from the headphones via Bluetooth and the TV audio via line in, mix both at a custom ratio, and send it back to the Bluetooth headphones. I know that it's possible to achieve such a setup using e.g. ardour and JACK, but I'm looking for an automated solution (e.g. write a script / small application that launches at boot time for setting up the audio devices). So how can I configure the mixing using only the command line or a custom Python/C/whatever application that uses some libraries? What's the best software stack to aim for? PulseAudio? JACK? Latency shouldn't be too much of an issue, as my grandmother probably isn't able to perceive lip sync anymore anyways, but of course it would be great to aim for the least delay possible. I'd love to hear some tips and pointers on how to approach this project. Thanks a lot in advance! Cheers, LuKaRo