Jesus Christ. Their product is straight out of every CIA/NSA/SecretGovernmentAgency movie of the last 20 years. All that 'magical' integration of news reports, signal and visual intelligence, financial records, military incidents, relationships and alliances, all kinds of intelligence, and making connections and performing analysis using a wide variety of really cool tools (that, judging by the videos, have incredible performance).
One can't help but think that the company started by some guy who was watching a bunch of spy movies and said "why don't we build that product and sell it to the government?"
When I interviewed with them and friends would ask what the company did, I would say "you know how in 24 Jack Bauer would call up CTU and say a name, then they'd find out everything about that person instantly, from their address to the last person they ate lunch with? Palantir does that for real.
Really impressive product, and they honed their demo to be pretty jaw-dropping.
Well, they actually sell two products simultaneously. They have a flashy, pretty good, front end tool. But what they really pitch is their very large and super scalable integrated ontologically driven back end.
I don't know all the details, perhaps one of their employees on here can provide a better, non-marketing-spin (please no "we're building an intelligence operating system" nonsense again).
My understanding is, that when they deploy they build a pretty sophisticated ontology for the customer. It's not a comprehensive ontology, but one that's domain specific to that customer -- which actually has the potential to work really well. So for example, the FBI would get one that targets law enforcement (people, places, addresses, license plates, violations, prior convictions, gangs, etc.), while the Army might get one that's focused around military operations (people, places, rank, equipment, unit, etc.).
They then custom write Java software that connects one of the customer databases to theirs, and probably either mirrors it and links it all to some metadata, or if it's federated, generates a pile of metadata with pointers/links back to the original data. The meta-data itself can actually be very very comprehensive. For example, assume the original data is a news report, that report might have everything from tagged people and their attributes to photos of some of the people, videos, maps, whatever (actually I think they just store a list of entity IDs and then store the rest of the junk oriented around the entities, but it's not that important). But basically it's a custom ETL tool for each data source.
I believe it's in that code that they also tell the Palantir enterprise backend how to map individual fields from the incoming data to entity types in the ontology. Assume instead of news articles, you are connecting to a phone book. You have to map names to people, numbers to phone numbers and addresses to locations (or whatever they have it called in that particular instance of the ontology at that site), etc. If it's a yellow pages, you can map names to businesses instead of people, etc. The ontology itself is mutable, so they can decide post-facto that they would like to add a new attribute to a person "place of birth". So if their knowledge base has phonebooks AND birth certificates, when you go to inspect an entity, like a person, it'll retrieve the place of birth and build you a nice dossier of that person.
In cases where they're consuming unstructured data (the lions share of government reports), there's no fields to map to the ontology. So analysts have to sift through each report and do the mapping with the front-end tool while they are conducting analysis. I know of at least one site that's in the process of hiring out a bunch of low-paid data entry people simply to go through and do this tag-map operation on their reports.
If ahead of time, the reports have been run through an enterprise named entity extractor, they can leverage those to populate the knowledge base. But in practice, quality is low, and the named entity extractors tend not to do a good job of determining different subclasses of entities. For example, you'll get a giant list of people from a document, but no indication if a particular person is a scientist, a politician or a terrorist (and all three of those categories might be entity types in the ontology for that site). In addition, the key factor here is the relationships between entities and between entities and their attributes. And most named entity extractors do an even poorer job determining that than just finding the entities. So the default at most sites is to have shifts of analysts manually tagging and mapping documents.
On the front-end, when you search for a person in the little search box, you can not only get documents that name might appear in, but you can call up a dossier on them with all the little various attributes and other bits and pieces filled in for you. Imagine a police record like you see on TV, with the person's photos and other facts and figures and such -- except that it's generated dynamically based on all the tagged and mapped meta-data, and come be composed of data from several different sources at once, like a phone book, a DMB database and a million news reports.
I'd say not only spy movies. There is an obvious affection for Middle Earth as well. Even a corner office is called Rivendell! He should just legally change his name to Saruman and get it over with.
As i recall the company was actually mentioned in a book i read - Blank Spots on the Map: The Dark Geography of the Pentagon's Secret World by Trevor Paglen. so you are on the mark here.
I'm sure Jesus thanks you for reminding him of this product but probably wonders what your point was, you prayer seems to have wandered off somewhere ... ;0)>
I saw a demo of their stuff at Joel's stackoverflow.com tour in sf last year. They were a sponsor and were recruiting. I got a save the shire t-shirt, anyway, It's a really cool product. At least then they were going after defense and government markets primarily. It sounded like they had oracle prices since every installation is basically a custom setup and their system is infinitely customizable. I don't doubt the valuation if they get traction. I know the big financial guys are also a prime market for this kind of product.
I'll be the first to vouch for the fact that this is the most amazing place to work. Interesting technical work, huge impact on real problems in the world, great culture, talented coworkers, and stock options that mean something.
Palantir is known around the valley - at least, I've heard of it from other Googlers (some of which have spouses/friends that work there). It's mostly known for treating their employees very, very well. Like, Google-level well. They also happen to be working on really cool technical problems, which makes them a big draw for talented engineers in the valley.
"Pay well" is relative. I believe they target the 75th percentile in total compensation, i.e. 25% of the time, people that receive offers from Google would receive more money by working elsewhere. Perks are included in that, but they're pretty tiny, eg. the free food is valued at about $5k/year (it actually only costs about $2-3k/year, which is why Google does it). That's pretty decent, particularly considering that people that receive offers from Google tend to be a small, highly-valued slice of the total programmer pool. I make more than twice as much (including bonus + stock options) at Google than I did at my previous employer, though that's not really a fair comparison since that employer was my first job out of college and there was a startup I founded in between.
The other reason I say it's relative is because your market value after working for Google at a couple years is significantly more than when you joined. After all, you made it through their hiring process, and then you've had a couple years to learn all about their search algorithms and infrastructure and data-driven culture and tons of other stuff that makes you more valuable to prospective employers. But if you haven't gotten a promotion (and promotions can be kinda difficult at Google, because all your peers excel too), you're being paid maybe 6-7% more than when you started, barely above inflation. So your salary - which was probably significantly above other offers when you joined - is now probably significantly under what those companies would pay you now.
Curiously, barring people that didn't fit in with the culture and left in their first 6 months, I know very few people that ever leave for another job. Most of the coworkers I know that have left (and that I miss) went to found startups. One is in YC WFP10. Ex-Googler startups seem to be getting picked up a lot for $5-10M nowadays, which is of course a whole lot more than a Googler would ever make in salary.
Earle,
You posted the fermilabs mystery letter: full cryptanalysis, you last comment is drink you ovaltine. I am incredibly interested in where this fits into the code...f0be58f2fd63 6c79d2e493e6..I am trying to break a code that is very similar to this but have no experience breaking codes so I have been researching codes and how to break them, this is the closest I have come, but still would love to ask you some questions, if you have time. I dont want you to break it for me, I just need a little guidance, when thinking this thru.
I'm concerned, they've pretty much started to saturate their target vertical in the government (military/intel). I've also noticed that new installs are starting to go in at pretty large discounts over a year ago. The government, while large, is finite in size. It doesn't take too many enterprise deals to max it out.
If they aren't making money yet, and at their tremendous burn rate, is another $90million really going to bring them to self-sustainment? I'd bet that the curve on their growth rate is starting to drop off now, the next logical place for them to push into is local law enforcement. But at their prices, compared to the de facto competitor (i2) I can't see them getting much traction except for maybe the very largest municipalities (I know they have one in the DHS funded NYC Fusion Center for example).
Their going to have to put more resources behind their non gov/intel verticals I think for this to make sense.
I agree, it's actually where they were born -- financial fraud detection. Their finance product is really mature and quite powerful. Not sure how big the market for it is though. They'll probably do well in other types of research areas, think tanks, large news agencies, etc.
They'll definitely have to start to push out into non-US governments though to survive. But from my experience, just due to funding differences, other governments are far more stingy with their purchases.
I don't know. All of the demonstrations I've ever seen have been forensic (after the fact) investigations. They spent a bit of time at the last conference trying to build a case for why their government product would be a great network security tool, but I never saw or heard anything that would indicate it could be used as a real-time tool.
Their focus recently seems to be on scale, lots of spin on how they can cluster and use billions of records on commodity hardware, link together installs at different sites, etc. etc. They did a demonstration of and investigation into the housing crash by analyzing millions of mortgage applications, then built aggregate bar charts and heat maps they could drill into a bit. It was "ok".
I think if they are serious about penetrating the network security market, they'll need all new technology that can be used in real time. But they seem to already be well regarded in financial fraud as a post event investigation tool.
Their initial installs are quite expensive. Small orgs <50 have been getting prices from $250k-$750k depending. I've seen larger organizations with $5-6million per install prices, but there are fewer of those. I've heard that their quoted maintenance fees are a bit higher than that.
I think at their last conference they said they had ~200 staff and still growing. So their burn rate is ~$30-$40 million a year.
This latest round will keep them alive for another year or two.
There is no publicist, no sales or marketing team and Karp adamantly believes that there will never be one. He says he is perfectly content to let word of mouth drive his business, in press and in sales.
It means they are completely rewriting the rules for selling enterprise software.
This is strategic spin -- an emphasis on having technically-savvy people fill sales roles under the 'Business Development' label.
About half the titles advertised as open positions on Palantir's website, across all locations, fall under the 'Business Development' category. They're requesting engineer-level technical aptitude, sure, but the work is lots of pre-sales customer contact.
Ok, that sounds more like an incremental improvement than a complete rewriting of the rules. I wonder what percentage of their total effort goes into that as opposed to building the product.
It's more like an increase in efficiency. The turn all of the engineers into what are essentially BD and sales staff. It saves them from having to deal with high-commissioned sales people, while enjoying the benefits of low salaried, just out of grad school, smart engineer types that clean up well (the company branded track jackets and massive quantities of hair product on display at any of the conferences helps too).
Folks, their customers are intelligence agencies and the government. They spread through referrals from the CIA to NSA to State to DOD to NIH to whatever the hell. Then they get referrals to their counterparts at friendly governments. The hard part was getting their first sale. Hell, I bet they're still not on the GSA schedule, even with tens of millions in government revenue.
Palantir Finance sells to hedge funds and financial institutions, but the whisper referrals work in that insular industry just the same.
These guys need salespeople and marketing folks like I need a hole in the head.
I used to work at a company that had basically one client, the US Navy, and it would still go to expos to show of its products to different parts of the Navy.
The government is not a monolith, where everyone shares the same hallway (although the Pentagon might come close...).
That I vacuum the floor in our workspace doesn't mean my startup has a janitor.
There is a well-established game around enterprise sales. It involves golf, steak, strip clubs, and (as tptacek memorably put it) gnawing off your own toes. If these guys are succeeding at selling big-box enterprise software while refusing to play it, then they are rewriting the rules and others will follow. I don't want to be over-optimistic about this, but it sure would be a good thing, and I find it more interesting than either their analytics platform or their bubble machine.
FWIW I'm the cleaner, kiln technician, website team, chief scrubber and bottle washer, etc., for a company too.
I realise they're doing things differently. If people come to them all ready set to buy they still have to go through some process even if that process is only payment - someone manages that in some way. Maybe they take online payment and do delivery with no human interactions from the company side but someone still set that up, they still have a sales person.
What an amazing line of BS. Anyone who's ever dealt with them deals with sales people. Please don't buy into this, they are by far the most marketing and sales oriented companies I've ever seen in 15 years in the government/military intel space.
It sounds like you have direct experience with them. If so, would you please describe what you personally saw or heard, that sheds light on this question?
If it's BS, perhaps it's BS designed to hoodwink engineers. It certainly got me excited thinking that it might be true.
My name is Ari and I'm a senior software engineer at Palantir. I was the second member of the Palantir Government backend team so I've been with the company almost five years. I've seen this place grow up around me. To describe me a bullish about this place would be an understatement.
I can categorically confirm that this is not BS: we don't have sales people, we don't have marketing people. This doesn't mean that we don't have people performing those functions -- of course we do! However, the people performing those functions are all technical folk who have stepped into the breach to do the needed work to grow our business. As a commenter in the TechCrunch article mentioned, "I'm the founder at my startup. Just because I sweep the floor sometime doesn't mean I'm the janitor."
I think what we said got a little garbled as it was distilled into the TechCrunch article. When we said, "we have no sales people", what was meant was that we don't have people who know nothing about the product, worrying about their commission and their quota and adding essentially zero value to the process. The people who come up with our marketing messages are not people with marketing degrees -- they're engineers who intimately understand our technology and the benefits that it brings to our customers.
Even I've stepped up into that role: along with writing our backend data importer and authoring our XML formats and processing pipelines, I also run our college recruiting program, and write for and edit our tech blog . Does that make me a recruiter or a blogger? Not in the sense of the word that most people would use it. I'm an engineer, proud of it and my technical achievements, and anyone who calls me otherwise better do it when I can't hear them. I'm an engineer that does recruiting and blogging, just like we have other engineers that do sales and marketing.
I look back over my time here and I marvel at the amazing place that we've built. It's an engineer's fantasy land that really exists - I can only think that this is what Google was like, circa 2000.
But don't just take my word for it, come and check us out for yourself. Shoot me an email to ari@palantir.com and I'll set up a visit, a proper interview, or whatever is appropriate for your situation.
I appreciate the response, but the impression at your customer's sites is that they are dealing with sales and marketing people. If somebody's job is majority spent on sales and marketing efforts, he's a saleman or a marketer even if he doesn't have a sales and marketing degree (degrees count for very little in the vast majority of sales and marketing positions in my experience). I actually have received business cards from people in Palantir who's titles (at least on the cards and in their resumes) are "Business Development" or something similar. Not "Engineer" with a bullet buried someplace on their responsibilities as an “account manager to XYZ agency”.
We've all filled many hats, I even take the trash out, or change the toner in the printer from time to time at my place, but that doesn't make me a trash man or the "toner guy", I understand what you are saying. But really, sometimes when a person's job is something other than their title or their degree would indicate, that's their job. Sorry, but if you spend 80% of your time engineering, you're an engineer. Likewise, if you spend 80% of your time selling you're a salesman. It doesn’t matter if your degree is “Ethnomusicology” and your title says “Forward Deployed Engineer”.
I know of a couple people there that have transitioned in the last year or so into full-time sales roles but started as engineers. That's not to say that they won't transition back to engineering later on. But for the time being, they are "salesmen". If you split your time more evenly, and it sounds like you personally do, perhaps you just fill several roles? I bet you have at least one person on staff who spends 3/4 of his/her time recruiting. But that's unlike some of your colleagues who, while they may do more than one thing (which not unusual at any organization no matter how large or small), spend most of their time doing sales and marketing. In fact I've just had people in two organizations who are deployed in theater right now relay to me how they purchased your solution after meeting with teams of sales people over the last few months.
There's no doubt that you guys are building some very good technology (deployment issues notwithstanding). But that technology has to stand for itself if what you and Karp are saying is to have any credibility. And I know that carpet bombing D.C. with advertisements, business development associates and sales people is not accurately reflecting your statements no matter how it’s spun. In fact, by making a big deal out of "not having a sales or marketing department", isn't that yet just another form of advertisement?
Statements like, "I can categorically confirm that this is not BS: we don't have sales people, we don't have marketing people. This doesn't mean that we don't have people performing those functions -- of course we do! However, the people performing those functions are all technical folk who have stepped into the breach to do the needed work to grow our business." are just splitting hairs and come off as untrue when interfacing with the sales, BD and marketing staff of your company. I’m sorry, but I and many people in the IC are immune to these types of Jedi mind tricks – and I’m giving you an honest assessment of what we all talk about in reference to your company. Please take this as constructive criticism. Efforts to not have a sales and marketing department, purely to obfuscate the presence of sales and marketing staff, and then pump up the lack of sales and marketing people, does not help the credibility of these kinds of statements.
I think something like, "what was meant was that we don't have people who know nothing about the product, worrying about their commission and their quota and adding essentially zero value to the process. " is fundamentally stronger to your message. "Our sales and marketing staff are also all engineers on our products" is much closer to the truth than what was presented in the Tech Crunch article and here. If that was the message that was getting out across the IC, I think that your company’s credibility would be much stronger. But right now it’s seen as precisely the opposite.
Here’s an exact quote from an analyst downrange I work with from time to time in reference to issues they’ve been having with their Palantir deployment, “they ultimately fell for a slick talking salesman that showed how well Palantir handles canned data. They got EXACTLY what they paid for.” Another from PM I work with from time to time, “do you know of anybody who’s happy with the Palantir install?” I’m not sharing these to knock your product, but to help provide you with some insight into where the minds of your customers are. And by customers, I don’t mean the acquisition folks and deputy directors who sign off on the purchase, I’m talking about the people who you are trying to help, the end analysts.
I know that in general, the people I’ve met from your company are extremely bullish on how you guys are doing, and that’s great. Their enthusiasm for what they are doing is commendable. But what I, and my colleagues in the community are not seeing is a steady, reserved hand of reality coming in and helping you guys turn the reality distortion field from 11 to a nice, listenable 5. Obviously you guys have to trumpet your product and make it look as good as possible, I get that. But don’t be deaf to the cries of your customers! One thing that I definitely came away with from working with a few of your deployed engineers is that they definitely want to do what they can to help the customer. But that implies not just listening to how you can be better, but also where you are falling down.
> "I've even I've stepped up into that role: along with writing our backend data importer and authoring our XML formats and processing pipelines, I also run our college recruiting program, and write for and edit our tech blog . Does that make me a recruiter or a blogger? "
Yes. It does. Depending on context or time spent on those activities. To give you an example at my own company, I'm an engineer, an application manager, a product manager, a salesman, business development, trainer, analyst, and half a dozen other hats. If I'm in doing a sales pitch, you can guess sure as the sky is blue that I'm a "saleman" for that context, I'm not an NLP algorithm specialist, or an Operations Manager, or a "toner guy". But it's certainly extremely helpful that I know those roles so I can answer those questions in the pitch if they come up. My education is in Computer Science, Mathematics and Management, it doesn’t really matter though if for that day, the job I’m doing is “Analyst”. Two years ago, I spent 80% of my time in sales, I was a “salesman”. I’ve now starting moving much of that work off of my plate and maybe only spend 40% of my time doing that. But to say that we’ve never had a “salesman” would be fallacious.
Outside of my own, or my colleagues experience with them, they literally wallpapered the metro station at the Pentagon with adverts. I've never seen that anywhere else in the system - ever. We're talking one of the most competitive advertising stations in the system, with many multi-billion dollar companies competing for the eyes of the senior military decision makers, and it's Palantir Tech that has everything from the floor to the ceiling completely covered, not Lockheed Martin, Raytheon, Northrup Grumman, etc. I refuse to believe that an engineer in a back room, one day, decided he was tired of hashing out java code and picked up the phone and rang WMATA and bought a quarter million in advertising. No, that was done by their sales and marketing staff.
Their last conference had ~700 attendees, fully 100 of the attendees were Palantir employees flown out from the home office to fill out the space and make a big presence felt. That's fully 50% of the staff at the time. Oh, and to helpfully answer questions I suppose, but very few were able to coherently answer questions beyond a superficial level. Me and some of my colleagues kept getting passed off to probably the same half dozen engineers over and over again -- sometimes humorously so ("didn't we just meet upstairs at the breakout and talked about something completely different?"). It also doesn't help when there's so many Palantir people hanging out in reception area that it's impossible to walk a straight line from one part of the conference to the next because of the clusters of track jacketed employees lounging about and giggling excitedly whenever a uniformed Army officer walks by.
And let's not forget the conference itself, two floors in the Ritz Carlton, with catered lunch of Filet Mignon and similar fare for 1000 people? With an open bar after the conference? (Thanks for the coffee cups guys! They're actually pretty nice!) That's not a DBA running that.
Some other colleagues of mine at a different site described a team of "slick haired, googly dressed, sales guys" (actual quote) descending on their facility to hawk their wares. The brass was impressed that "even the sales guys knew the technical stuff". Which of course happens when your sales guys have engineering degrees -- I'd actually like to see more of this in the future in general, so it's a trend I like.
Direct personal experience? For every decent engineer of theirs I find at a site, I find 2 or 3 that don't seem to do anything except hand out stock business cards with direction to go check out operation tradestop or whatever their online demo is called.
If that doesn't persuade you that they have invested large dollars in a sales and marketing apperatus, my own interactions have been through people with "Business Development" in their titles.
Yes my friend, just because they happened to have graduated from Stanford or CalTech or wherever with an Engineering degree does not prevent a person from being in sales and marketing.
Like most things that come from the company, there's actually a pretty interesting kernel at the core of a giant pile of BS. The fact that their sales staff actually are engineers by training is a great thing to see in a large company. I've been on the receiving end of too many pitches by clueless sales guys with marketing degrees, when what I needed in that hour I was able to allocate to hear the pitch, was some hard technical answers. Having guys who know their stuff getting out there and pound pavement is no doubt a big part of their rapid success. But let's not be mistaken, they are salesmen, and the company does spend large amounts of money on marketing, press releases, advertisements, etc.
They actually stand out as much for their product as they do for the fact that they're by far the most "present" company in their competitive space on billboards and marketing materials in the entire area.
Edit: with hindsight, it just occurred to me that if the original quote were true, they wouldn't need a $90M Series D in the first place. Filet mignon at the Ritz, on the other hand, adds up pretty quickly.
Yeah. While I think they have pretty good product, they have the reality distortion field turned up to 11. It permeates everybody who deals with them -- including at times their own staff.
The consensus among people I know is that they are paying gobs of money to the owners of these papers to get this kind of coverage. Why run a half page add when you can get the paper to just send a reporter out to your office to give you a favorable writeup with pictures of the CEO looking impossibly cool and not even bother to interview the end users of the software who might have something negative to say?
Well, it's easy to re-write the rules when you have a cash cow called the US government buying loads of your applications. You just don't need to chase after customers anymore. Right?
Does anyone else get the feeling this sort of software is probably sold for tons of money but rarely useful? I could be way off here but just from glancing at it, it looks like one of those things that is probably too complicated for the average government employee to use. So they end up having to hire outside consultants or something. It looks like a research tool that is one step removed from the lab.
Most of the people that will be using this tool are intel analysts. Most intel analysts used to be in the military at some point. To be an intel analyst in the military you have to score relatively high on a test called the ASVAB (basically an aptitude test). There's still lots of old guard in the field who can't work a mouse, but the recent crop of analysts were all trained and worked on systems of similar complexity and size (ASAS, JIOC, DCGS, etc.).
In other words, we're not talking about the typical government paper pusher.
That being said, it's still a complicated tool requiring the training and understanding of some pretty complicated concepts that most analysts don't know about or understand (like ontologies).
Given the sensitive nature of some of the data they work with and the fact that 70% of their business comes through the [US] government, I wonder if their engineers all have at least some level of security clearance. Or would palantir just as easily sell the same product to the US government as it would to another government?
Their "forward deployed engineers" need clearances to work on site. Palantir installations are an extremely lengthy process required a couple full-time people on-site. The places I've seen it, it's typically taken 6-12 months to get it "turned on".
I smell gov't opportunity for SAS to acquire Palantir, as they kinda-sorta are in the same industry (data mining and analytics). And, the way they treat their employees is very similar (i.e., good work-life balance, in-house healthcare and childcare, killer culture that is the envy of even Google):
Word on the street is that they are trying to position themselves as an Oracle acquisition. Their valuation seems a bit inflated, but I could see them selling for $300-500million.
Depends on the kind of relationship you're talking about, but generally speaking, someone has to map the data from the original databases into Palantir entities/properties/links. But—and this is kind of a big deal—the Palantir product does not have a fixed set of entities/properties/links/etc. So if you're more interested in tracking, say, computer systems, networks, and packet traffic than terrorists and bank transactions, it's fairly straightforward to define a new ontology that only contains the things you care about.
https://devzone.palantirtech.com/display/pgdz/Integrating+St... has a pretty useful overview of how you can take raw data from an existing source (granted, this source is XML data, but it could just as easily come directly from a regular database) and map it into a Palantir stack.
So basically, the heavy lifting palantir does is in the presentation and data abstraction. With a person still needed to define mappings across data sources. Is this a correct statement?
With (seemingly) limited off the shelf mapping abilities I'm surprised that they are looking to sell more as an appliance, and less as a turn key service with consulting engagements etc... Neat product.
I should probably mention that I work at Palantir, but [insert boilerplate disclaimer here].
Yes, you definitely need a person to define the mappings across data sources. However, this is basically a one-time setup task; once the mappings have been established, new data from the existing databases can be continuously imported. I don't have any stats on how long this initial setup phase takes at real customer sites, and if I did I'm sure I couldn't share them publicly, but the goal is definitely for it to be something that doesn't require an entire team of consultants to deploy and babysit.
If you want to see the actual government product, and how it works, you can sign up at https://analyzethe.us/ for an account. This is a real live Palantir instance with real data from data.gov and it's open to the public.
At the dozen or sites I'm aware of, only a couple are where I'd say are in "full operational mode" after 7-9 months. I've known a couple that have been in a protracted (even by that standard) deployment stage for >12 months. Every place I've seen has a 2-4 Palantir engineers working F/T there during the deployment phase and at least 1 or 2 FTEs to maintain the ontology as mission requirements change ("oh, now we need VIN numbers associated with vehicles!").
I'd imagine it's getting faster as staff gets trained and various little ETL tools are built to funnel data into the backend. But it's been a very frustrating process for most of the customers to drop the money on the deployment and not have an operational tool for the better part of their first year, and the maintenance tail has been reported to be some percentage higher than the original purchase (that may have changed, I've noticed the pricing structure has been changing quite a bit over the last year).
Of course, I also understand that when the customer says "pump data source X into Palantir" and data source X requires a 6 month approval process to hook up to, it can just take a while.
They map data from different sources onto a common semantic ontology. From that, relationships can be derived. It works better with nicely structured data, but people can be leveraged to manually tag entities in documents and add those to the appropriate part of the ontology.
A very large part of their strategy is crowdsourcing named entity extraction. Basically, as you read a report, you hilight and tag all the entities of interest. It provides incredibly high quality entities and relationships on an enterprise ontology.
The problem is that it scales incredibly poorly. Imagine hilighting every person, phone number, location, etc. in a document. Then linking them all together manually. Then imagine having a few million documents you have to do that on. It's been an enormous problem at all the sites I've seen it deployed at.
When I was playing with it, it seemed like a really handy way for a human to sift through a lot of data.
Possibly that's where the human-based spin is coming from.
One can't help but think that the company started by some guy who was watching a bunch of spy movies and said "why don't we build that product and sell it to the government?"