Cognitive Services takes individual words and uses machine learning to piece them together into meaningful sentences. The SDK takes care of finding the microphone, sending the audio to Cognitive Services, and returning the results.
In the next recipe, we are going to use language understanding to determine the meaning of the speech. After that, we are going to make a smart bot using Bot Framework, which builds upon the language understanding to give state and logic to the ordering kiosk. You can use speech as an input to that system.
The Microsoft Speech SDK allows you to account for accents, pronunciations, and sound quality through its custom speech service. You can also use Docker containers for environments with limited connectivity to the internet.