How to formulate a voice instruction

[JerryMatera]

How do I formulate a good voice instruction from the osrm api response? When you query for the next instruction using this package you get typedef of the current instruction, the next instruction, and the distance. Current instructions include distance, instruction text,distance and instruction location. I want instructions like in Google Maps that factor in distance in the voice command.