flight-plan-fly

Neat! I have been thinking of trying to build something like this as a side project, but knew it would be difficult to get good results without having a team of aviation folks labeling data.

Do you think your model's word error rate would decrease if it had access to raw feeds from towers or was trained on data from even more airports?

Do you have any plans to make the data searchable/discoverable in real time? Something like being able to search for any audio transmissions regarding a specific flight by number, etc., available on ADS-B exchange, Flightradar, FlightAware, etc., both in text and audio form.