You can get LoRa concentrators in m.2 form factor, so it shouldn't be that difficult to implement with what they are announcing.