Palantir, like all too many firms creating innovative software, went to where the demand for innovative software was: preventing terrorist attacks.

New, large scale software never gets adopted in the private sector, because the suits that run them are too myopic to think technology can solve core issues that have never been solved with technology before. They only invest in improving the technology they already have in the segments of the business which are already automated.

Oracle sold its first relational database to the United States Air Force. Nobody else thought they needed it at the time.

Palantir actually offers a very general purpose toolset, but it is the intelligence agencies which were the first to have a giant pile of money and a problem that never goes away.