Went to a luncheon and sat with some IT Directors at a Fortune 20.

I asked what they were seeing and excited about.

They kept explaining that Foundry (Palantir's SaaS BI platform) is better than EVERY other alternative (and mind you, they've used every other major vendor as an F20). I kept asking what was special about it (Did it re-invent data models? Is it faster/cheaper than MSFT, GOOG, AWS, SNOW?)

I kept getting circular answers (advantages without addressing design consequences) until I realized (to myself) that what they were describing as "great" had nothing to do with the Palantir tech.

It was great because Palantir's sales people had taken a top down approach (getting CEO's blessing) and had the "green light" to greenfield data solutions and cut through internal bureaucracy/silos about connecting datasources to find revenues or savings. This is CEO (since fired) kept bragging to shareholders about rubbing elbows with Palantir's Alex Karp and gleaming with joy about the potential of their AI collaboration.

That's the impression I get about PLTR.

They're like if McKinsey was re-loaded with software, and sales engineers and they hunt C-Suite and government clients to "speak AI." I haven't looked recently, but one bearish sentiment was that they need growth to sustain their high P/E, and there are only so many more governments/CEOs in their addressable market to add.

It's literally just "better then what people had" + they're willing to work through government and military contracting processes so it can actually be deployed in those environments.

They have a lot of "forward deployed engineer" roles which basically means staff with security clearances who get locked in SCIFs and provide on-site technical support.

Which is really why they keep getting hired: when you write into your contract "it stays on premises and technical support can't take logs off site" they agree to it (at a hefty mark up because all of that sucks to do).