00:00This is a company pre-AI, before LLMs came to the scene, and they have done a lot of
00:06customized work for government entities, large defense companies, where it helps make sense of
00:14big data. Basically, it goes back to that big data days where if you're dealing with large amounts of
00:22data, they can help you create a pretty dashboard that can give you the insights from unstructured
00:28data. Now, with all the LLMs coming to the scene, they clearly have made a pivot towards incorporating
00:36LLMs as part of the solution. And what we have seen is accelerating growth over the last 13 quarters.
00:45So this is the 13th consecutive quarter of top-line acceleration. And 93% growth is very impressive
00:53for a software company. But when I look at the customer base, we're talking about 1,000 customers.
01:01So it's very high ASPs, very high monetization at their customer base, which, like I said,
01:08half of this is coming from the government. And so from that perspective, a typical enterprise
01:14company like Microsoft has got over 100,000 customers, even a Snowflake or Databricks has
01:21got 20,000 plus customers. Palantir is just catering to 1,000 customers. And that's the real risk for me
01:28when I look at the, you know, the growth trajectory next couple of years.
01:32So you're saying that's a source of fragility rather than a source of strength for Palantir.
01:36Yeah.
01:37What could disrupt that? I mean, does that make Palantir more vulnerable to a competitor coming in?
01:42Yeah. I mean, the fact that the CEO has been so vocal about Anthropic and OpenAI and how suddenly
01:49out of all the companies, Palantir is supporting open source. I mean, this is a company that has
01:55benefited greatly from selling a proprietary solution. And now it's coming out and being so
02:01vocal about open source support. So that, I think, gives you an indication that they do see the
02:08competition coming from the likes of OpenAI and Anthropic, as well as the hyperscalers. I mean,
02:14what Palantir did really well was they have this concept called FDE, Forward Deployed Engineers,
02:19that were part of their employee base. So when you're installing a solution, the FDEs actually
02:26help you get there faster. And now all of these companies, whether it's Microsoft, Amazon, they have
02:33this FDE team that they are using to deploy their own LLM. So I think competition-wise, I feel there
02:43will be more competition from Palantir and they can see that, which is why they are being so vocal with
02:49OpenAI support. And I mean, the growth trajectory, they guided to increase top line growth. So it was a
02:56beat and raise quarter. But to my mind, the growth rates will normalize. And that's where the higher
03:01valuation kind of creates a risk.
03:04You mentioned Anthropic, because we're looking for a potential IPO maybe later this year. Bloomberg News
03:10has got a story that Anthropic has struck a $10 billion deal for computing capacity from a months-old
03:17infrastructure startup, making the cloud maker's latest effort to keep pace with demand for its products.
03:23What is that all about?
03:25I mean, right now, Anthropic is just getting compute from wherever they can, whether it's SpaceX or Meta,
03:32or in this case, this small company, Bitcoin mining company that has compute to sell.
03:38Hold on, hold on. A Bitcoin mining company is now providing compute for an AI play.
03:44Yeah, because at the end, it's a data center.
03:46That's the dream for these Bitcoin mining companies once they make the transition.
03:49I mean, compute is agnostic of where, you know, that data center is set up, what purpose it was used
03:59for in the past. It doesn't matter. Compute is unique processing capability with life power
04:05connected to a grid. That's all you care about.
04:08And just to be clear, we never used the term compute before, did we?
04:11Not in that way. We used it as a verb.
04:13Yes. We would never set as a noun.
04:15Yeah.
04:16Right. See, I just want to make sure I'm calling these tech people out on some of this stuff.
04:20Well, they make up a lot of words.
04:21They do.
04:22You know, I need more compute.
04:24Is that what you?
04:25That's what all these OMs need.
Comments