Leaked Palantir Doc Reveals Uses, Specific Functions and Key Clients
techcrunch.com
techcrunch.com
----------
Palantir is trying to raise another ~400mm to finance their operations. And I honestly think this just another PR stunt to attract investors. The 'secretive' attitude is just a game of deception. They are deceiving investors into believing that the company has a real product. However if you talk to any FDE or Embedded Analyst, you'll immediately realize that this is a consulting company. Which should be valued at far lower multiple than a real product company.
Every new contract requires dozen extra people...That's why they are on a hire spree right now, (because of a large 'consulting' contract with a an oil company, that should stay unamed). However, with oil trading at 50s, they are scared and they trying to get some cash to cushion a possible contract cancellation.
To be continued...
I've heard this as well. They essentially hire customer-facing engineers to go out and scope specific contracts and literally implement custom solutions with small generic software frameworks they developed in house.
On the other hand, the customer has results after 6-8 weeks. How much software can really be built in that sort of time frame? Not much...
In my opinion Palantir is a product company, but the product is deeply integrated in to an enterprise. And doing that takes time and people.
I don't see how you can simplify that into calling it a "dev framework".
Fair enough. But you may be underestimating the value of production-quality fast-development dev frameworks in enterprise.
Remember, these are guys who are used to seeing 5-year project timelines for anything that touches more than one system. (And every business case touches more than one system...)
One other thing to note is that while they build these custom solutions, the second you touch their stuff that support policy goes out the window, not unlike what Oracle does.
EDIT: n/m @dang answered this indirectly a couple of comments below.
NSA as a customer has been widely reported in popular press [1].
[1] http://www.forbes.com/sites/andygreenberg/2013/08/14/agent-o...
A: No, the CIA is a front for Palantir.
- Peter Thiel
edit: and I'm a Libertarian, so I'm biased against reaching this conclusion based on my own observations.
That was my reading as well - Techcrunch essentially reprinted marketing materials as incontrovertible fact for this story (continuing a long tradition of similar journalistic excellence). One particularly troublesome sentence:
"...[Palantir] is capable of building comprehensive models of activity to detect suspicious anomalies and is even able to provide immunity to fraud thanks to strategies the founders learned while still at PayPal."
Well there's an objective, accurate statement. Even Paypal itself is far from "immune to fraud". Palantir appears to be doing some interesting things, but this article is ridiculous.
It is more of an "under moderation" thing than a "dead" thing. It's on our list to make that explicit and to provide a mechanism for the community to make most of the moderation decisions.
To be fair, this description makes it sound like an ETL tool, just like Ab Initio, Datastage or Teradata, where the companies involved all make serious amounts of money from consulting (and the server licenses don't come cheap either!).
These tools can be pretty powerful if set up and programmed correctly; it's definitely not as simple as pointing them at a bunch of data sources and expecting magic to happen.
If Palantir haven't found a magic way to integrate data sources without any configuration, it does sound rather deceptive.
That said, if they manage to get systems up and running without 30-stories-high stacks of never-to-be-read paper, that'd be a revolution in itself.
I'm not sure I can think of many software companies with that kind of startup investment.
WP says they have about 1,200 employees, or a burn rate of about $300m/yr. Those figures are old, so this new round must be for another year of operations.
I believe I had heard in a previous discussion that they were planning on going public in 2015, so this is a bit of a surprise.
Last I heard they were still also a Java client application for the analytics. They appear to have had an HTML5 client in development several years ago, but that doesn't appear to be the current product (even videos as new as a month old still use the old client which they had up as a free demo for years on their site). [1][2]
So is this just for continued expansion or is Palantir some kind of hi-tech charity case people prop with more money than sense continue to prop up? My guess is the former, but there's some arguments it's really the second.
I work with Palantir guys and helping them with data sources at a large banking firm.
Everything is custom, needs to be customized and then it needs to be changed every month.
The magic bullet here is hard work and smart people.
Palantir hires young smart (but not street smart) Ivy league engineers that have no clue how actual enterprise software works ... much like the founders and management team of Palantir.
Not always. A lot of software companies that provide enterprise-level software but aren't the SAPs and Oracles of the world often supply in-house implementors, who work for weeks or months with customers to integrate those products into existing system (at a cost).
Every new contract requires dozen extra people...That's why they are on a hire spree right now, (because of a large 'consulting' contract with a an oil company, that should stay unamed). However, with oil trading at 50s, they are scared and they trying to get some cash to cushion a possible contract cancellation.
To be continued...
> "It’s the combination of every analytical tool you could ever dream of."
> The Pentagon used the software to track patterns in roadside bomb deployment and was able to conclude that garage-door openers were being used as remote detonators.
> With Palantir, the Marines are now able to upload DNA samples from remote locations and tap into information gathered from years of collecting fingerprints and DNA evidence
> The leaked document cites a 2012 study where 96% of the surveyed war fighters in Afghanistan recommended Palantir.
Useful, but nothing nefarious in itself
Insiders are selling shares to private wealth management firms / asset management firms. It's amazing how much hype can do in the private finance world when you have the perception of momentum and herd mentality. Look at Clinkle raising $30mm ... this is the same thing happening at a bigger scale. Look at Groupon... same thing.
Here's one to S F Sentry Securities, Inc. http://www.sec.gov/Archives/edgar/data/1321655/0001321655140...
Morgan Stanley took over $25mm in fees to arrange these sales... the reason for that is because they know that there is no way Palantir is going to have an IPO in the near future -- or else they would be salivating to do this kind of transaction for free to gain favor with management.
The fundamental problem with Palantir is that they are a government contractor masquerading as a product company. $200mm in government contracts comes with enough overhead that there is not a ton of margin. You can see tons of defense contractors in the midwest doing hundreds of millions in revenue.
Take a look at their revenue numbers for federal government (which is likely a major portion of their ~$420mm in revenue): http://goo.gl/v7JzT5 http://goo.gl/AzZeuR
http://www.sec.gov/Archives/edgar/data/1321655/0001321655140...
"Delaying" the filing of required disclosures makes it more difficult for investors and potential investors to make informed decisions, and possibly defeats the purpose of requiring its disclosure in the first place.
https://www.youtube.com/watch?v=aJ-u7yDwC6g
Here is an interesting video from them talking about data analysis of Iranian arms deals used to show off some of their tools used by the intelligence community
If the analysis capabilities are so great (and they appear to at least be good), then there is likely tremendously more good that can be achieved by opening up the code to the software so that anyone can use it to tackle local and global problems around the planet. Instead, the company is using it to arm secret police organizations and financial organizations.
They get paid handsomely.
> more good that can be achieved by opening up the code to the software so that anyone can use it to tackle local and global problems around the planet
This attitude bothers me. Palantir is not a charity. Why are you holding them to these standards?
>This attitude bothers me. Palantir is not a charity. Why are you holding them to these standards?
I fail to see how that statement is an 'attitude'. I'm also not holding them to any standard in that statement. As far as I can tell, it's only a statement of fact. Or at least, perceived fact.
Do you think the statement is not accurate?
And yes, it's an attitude/opinion though it sounds like you're just nitpicking words. I'm not the OP you responded to, but it's clear your attitude/opinion towards this topic is that this software should be given away for free, with the sole reason for that being that it would "[do] more good".
They invented it, they didn't steal it (presumably), and they're choosing to use that product of their intellect in a way that makes them profit. Heck, you could argue that they're already doing good by making it available to law-enforcers while still profiting. In which case your entire point is that they're not doing enough good according to your opinion.
The statement itself is too vague to be accurate or inaccurate.
Look online for people with DCGS on their profiles, you'll be able to trace it all back to trying to reproduce Palantir down to the web GUI. Some decent journalist should be able to piece a lot together if they just trolled LinkedIn. A lot of it is unclassified but technically FOUO so not quite cleared for public consumption. The reason I figure it hasn't been researched thoroughly is because it's extremely boring DC area "tech" news that is more about topics that are of little interest to the tech community and is boring even by DC tech standards. Almost every company here is some Big Data bullshit company that's writing something about 8 - 14 years behind private industry timelines from an engineering standpoint and held back by politics, funding, and sheer disorganization endemic to enterprises but especially bad in the bloated Pentagon budget.
And the actual codfiying isnt hard it undersatning the domain problem and how to apply the verious tools.
Having engineers that will want to work in this area and can pass the citizenship requirements is harder back in the day in the UK noprmaly all 4 gradparents had to be citizens from birth for DV (TS)
So in short, good engineers with ineffective domain knowledge might make a great solution... that sadly doesn't solve what the person with the check wants up-front but you can put the most capable domain experts and managers to lead a team of bad engineers and the software will be both expensive to maintain (raising prices for the unfortunate customers) and perhaps just plain not work very often, causing more band-aid "solutions" that riddle cash-strapped bureaucracies and pass the dysfunction onto us as customers. I'm confident that tech is not the solution for most dysfunctional organizations and that IT projects just bring out the worst in them.
BTW I was "banned" from writing any code that would go into production as a condition of my director being allowed to employ me.-
Well they apologised, I guess. No resignations, no charges, no referring evidence to police. But yeah, they did apologise for it, so they got that going for them.
I pitched in front of some people from ManTech once and knowing that they acquired HBGary made me somewhat uncomfortable - and made some negative things I heard about ManTech all the more believable.
The perks are nice. They have an onsite chef so you can have a catered meal for breakfast, lunch, and dinner. Apparently a masseuse comes in every month(?) or so for free massages. Seems like a nice gig for the programmers.
That's a bad deal. Ask for enough money to hire your own chef and masseuse instead.
If I could go into work and not have to worry about preparing a good, healthy meal for myself (or drive somewhere to get one), that would be great.
You'll actually start to curse the catered food part if you want to actually conform to some sort of diet, weight loss or otherwise. The food can create a culture of longer working hours, so leaving earlier so you can go make your own dinner will create career challenges later on. SV is a large suburban place, not catering lunch just wastes a shit ton of time for the company itself and creates a lot of traffic.
And the masseuse service isn't that expensive for the company to hire once a week.
They have a chef available if you want your meal prepared there--no one's forcing anyone to eat at work. Also, from what my friend tells me it's not like one thing is being served each day. You have many options and they cook what you order. Seeing that he's seriously slimmed down while eating most of his meals there, I don't think his diet is suffering at Palantir due to the catering.
I don't really get what your point is other than to split hairs over how "nice" the perks are.
EDIT: Perhaps this quote from the website will explain the catering: "At Palantir, you will never be hungry. Chefs craft weekly menus consisting of organic and local foods, and we serve three meals a day. Family, friends, and guests are always welcome, and to-go boxes are always available."
All of this indicates that there's no big mystery here - they built a relationship-analysis tool that doesn't suck horribly. When you're selling to banks and the government, "actually works as promised" is rare enough to make you a billion dollar company.
Palantir is more hush-hush about some of their work, though this seems to be more 'competitive advantage' and less 'military secrets'. New initiatives and project restructuring are kept fairly quiet, and their customer lists are obviously secret, but none of this is shocking.
TechCrunch seems to have discovered what looks like a 'secret' press release about how great Palantir is, and it's being spun as dark secrets instead of info that any Palantir employee or tech talk is happy to divulge.
I personally don't agree with much of what he says, but still gotta give him props for his business savviness.
In some ways he's better than Musk. But the projects Musk has choose are far more interesting and engaging.
This person drank a bit too much of the valley kool-aid already, it seems.
The only question I have is how could investors be this dumb?
I have heard that some investors in the prior round (all of what, 12-18 months ago?) are trying to get some/all their money out after seeing the curtain pulled back. It has to be somewhat of a stampede for the door gaining momentum.
Still, while it is a cushy place to work (smart people, nice perks, okay money and stock that is supposedly valuable on paper until everyone rushes for the door), I also don't understand why smart people work there - unless they don't.
Maybe the smart employees are the ones who left 2-3 years ago.
Maybe in light of growing inequality and youth unemployment marijuana is seen as one of the diversion tools to prevent the poor from taking political action. Especially because it has roughly the intended effects, when used regularly. Note that it is not entirely out of the question that people in power think that way, for example the 60-70's were perceived as a "Crisis of Democracy" in parts of the establishment.
That's just nine months after the initial release of Vista. So hardly surprising to see XP on a desktop.
Current day job is performing service IT for a security-paranoid bank that nevertheless runs XP exclusively internally, so I guess I'm just hypersensitive :D