This is the first fallacy of software engineering. In general, nobody cares what the tech stack powering a solution to a given problem is. If it solves the problem, then it is considered a success by customers. There are exceptions to this rule, but there aren't many, and they don't really apply to Palantir's model. In their case, as long as the system spits out the data it is supposed to, they could have used Visual Basic to program the backend and customers would still buy it if it worked as advertised.