The company now has four hundred engineers and a machine-learning engine that is marching ever closer to Airbnb’s holy grail: the nimbleness to extract from perhaps ten thousand available listings for certain dates in, say, Paris, only the five or six that each particular user would like most.
