Pro-code capabilities like Compute Modules, OSDK, Jupyter, and Custom Widgets are putting the final nail in the competition’s coffin. There was a slight air gap earlier for teams that wanted to work directly with code repositories. AI-FDE has now smashed that gap on all fronts,

World remembers the brave, not the agreeable!

This won’t happen in companies that use Palantir Foundry/AIP, just a click away to see entire lineage. $PLTR 🤷‍♂️

We implemented an integration layer on widely used enterprise software using the Palantir Foundry/AIP suite. It outperformed their own agents and AI integrations by an order of magnitude (Accuracy, Ease of use, Scale, Robustness and many other metrics). That’s the power of

Everybody wanted to serve $PLTR ontology in their own way but don’t want to admit it 🤷‍♂️

For enterprise-grade applications, Claude is performing far better than other models, at least for the use case I’m building. Claude is clearly ahead, with GPT a distant second. The rest are trash. The bottom line is that I don’t care which model performs best. In Palantir

With all the buzz around major agentic AI players, most enterprise customers are experimenting heavily, but almost everything is dying in the PoC phase. The next phase will be about discovering who can actually deliver. That’s when a massive wave of customers will realize

I’ve been reading several pieces about context graphs, and many of them oddly drag Palantir’s Ontology into the discussion, often based on incorrect assumptions. Some claim Ontology is static while context graphs are dynamic. Some argue that only FDEs can build it. Others say it

I now see what Chad has been saying and quietly pushing over the past few months. The fast pipelines are blazingly fast. I’m not sure what’s behind it - Polars, DuckDB, or something else but a pipeline that used to take 15 minutes now runs in a few seconds. Sometimes I have to

Whatever agent context problems the industry is wrestling with, Palantir figured out long ago. $PLTR 🚀

This is called HyperAuto, built by Palantir and working in production for ages. $PLTR 👏

More things cooking on this front we’ll share very very shortly 👀 Things are about to go crazy

For many, conviction rises and falls with the stock price. Be the master of your mind, not its slave.

It’s visible in plain sight: Palantir is the only company actually making money from AI. Everyone else is running at a deficit. $PLTR 👊

The secret to finding great stocks: Stay within your circle of competence — especially in the field you work in. You can spot what truly works, even if it’s not attractively priced yet. Wait. When the loudest voice in the room — with zero expertise in that field — starts

Recently, I attended an exclusive summit hosted by a product company, where I got to see many applications that large enterprises and consulting firms are building. I can confidently say nobody is coming for Palantir’s lunch in the foreseeable future. $PLTR 👏 At the end, they

Steak dinners only take you so far. Eventually, execution drives everything. The Burrys of the world can call it consulting or whatever they want. Just sit down and execute one real enterprise-level AI use case, you’ll quickly understand where the industry actually stands. I