Inside the founder factory known as Palantir, America’s most polarizing company

Chances are, you have an opinion about Palantir.

“With any person, company, or concept, the general public really only has space in their head for one characteristic of it,” says Palantir alum Marc Frankel, cofounder, board member, and former CEO of Manifest, which creates software and AI “bill of materials”—think ingredient labels for critical software. “Biden: old. AI: scary. Palantir: secretive.”

Frankel worked at Palantir from 2013 to 2018, and whether the one idea in your mind about Palantir is secretive or something else, it likely exists somewhere in this band of public opinion from the past year.

Believers: Palantir’s a “category of one” company, according to Everest Group partner Abhishek Singh in a blog post last year, crediting its forward deployed engineering model where it embeds teams with customers to tailor its products to their business.

Critics: Conservative comedian Tim Dillon calls it a “shadowy military-CIA contractor” building a “digital prison.”

Investing bulls: “They’re the best software company,” concluded Gil Luria, head of technology research at the financial services company D.A. Davidson, after Palantir’s successful Q3 earnings report, where it closed 204 deals worth $1 million or more in those 90 days, 53 of which were worth more than $10 million as companies flock to build on top of Palantir’s Foundry and AI platforms.

Investing bears: “I just don’t know how this company ever grows into its valuation,” said Dan Nathan, a former trader turned financial media personality on CNBC and podcasts, referring to Palantir’s market cap hovering around $400 billion, or 100 times revenue.  

As Frankel adds, whatever your one thing may be, “it just becomes this trope.”

Are you thinking about your feelings about Palantir right now? Good.

Now it’s time to add another idea about Palantir, no matter your beliefs. This is a story about what really underpins Palantir’s success. It’s not its products. It’s not CEO Alex Karp or its other high-profile cofounders.

The idea is Palantir: unparalleled talent magnet.

• How has Palantir attracted such an astonishing array of talent?
• How does the company get so much out of its employees? 
• How have hundreds of Palantir employees gone on to start their own companies? 
• What can any company that wants to build this kind of talent density and financial success do to emulate Palantir?    

---

If your company envies Palantir’s success—financial, cultural, or otherwise—the story of its employees turned founders reveals:

• how to hire the Palantir way;
• how to build a dynamic workplace culture that delivers measurable results;
• why traditional corporate structures can be impediments;
• the ultimate secret behind the company’s success.

If you admire Palantir’s mission, this Premium story offers:

• our exclusive list of 315 former employees turned company builders;
• how these founders are advancing and adapting what they learned at Palantir for a new generation of businesses.

If you care about Palantir as an investment, this story will give you:

• a new way of looking at PLTR and what to look for when considering its future prospects;
• ideas for both private and public-market investments via the comprehensive list of Palantir alumni-led companies.

If you’re not a Palantir fan (to understate how many of its critics feel), this story explains:

• the real sources of the company’s strength;
• exactly how and where the company’s influence is spreading.  

Signing up for a mission

As recently as a decade ago, Palantir was largely unknown. It had offices in Silicon Valley and New York, not unlike Google or the other hot tech companies it competed with for talent, but it wasn’t on most people’s radar.

So how did it attract people to work for it?

In a word, mission.

Multiple former Palantirians eagerly volunteer how the company’s mission made them want to work there. “There were different articulations over time,” says Cobi Blumenfeld-Gantz, who worked at Palantir from 2014 to 2020 and then cofounded Chapter, which uses AI to help American seniors find the optimal Medicare plan at the lowest cost. “The one that stuck the longest was ‘Palantir solves the world’s most important problems at the world’s most important institutions’ … which is the most amorphous mission but it’s great. It’s exciting.”

“For any precocious undergrad, that sounds really appealing,” says Howard Zuo, cofounder and CEO of Dataland, which builds AI agents for complex operations and customer support. He did summer internships at Palantir in 2015 and 2016 and then worked there full time from 2017 to 2020 before pursuing his first startup. 

Ty Wang, CEO of the AI-native healthcare benefits platform Angle Health, came to Palantir after working for various U.S. government agencies thanks to his Stokes Scholarship. (Wang received a full college scholarship as a STEM major in exchange for work inside the Department of Defense and other agencies.) “The beauty of Palantir’s culture is that everyone had probably different reasons to be there and cared about different kinds of missions,” he says, “but you had the opportunity to work on the things that you really cared about and have tangible impact on real-world outcomes.”

Perhaps no one articulates the breadth of how one could find meaning in Palantir’s mission better than Frankel, who was 27 when he started at Palantir in 2013, coming from working in consulting. “It was absolutely ‘make work,’” he says of his experience. “Dig a hole and fill it in.”

By contrast, at Palantir, with no background in financial investigations, he could support finding inside trading because of Palantir’s software. With no background in counterintelligence, he could support counterintelligence work. With no background in fraud, he could help teams discovering a quarter-billion dollars’ worth of mortgage fraud.

“I worked on an investigation of a homicide that was a wrong in the world that needed to be made right,” Frankel says. “I can’t do push-ups. I’m never going to rappel out of a helicopter. This was a chance for me to feel like SEAL Team Six. This was my superhero cape.”

Like every tech company, Palantir offered both salary and stock (though as others tell me, often below market rates because you were signing up for the cause). “It really paid me in mission, and that was what was most appealing to me,” Frankel says. 

---

Luba Lesiva—who worked at Palantir as its head of investor relations from 2014 to 2016 and is now the sole general partner at Palumni VC, a firm that’s focused exclusively on backing Palantir-led startups—currently has a list of “379 companies that are founded or led by a Palantir alum that are still active and private,” she says. Some of them are so nascent, Lesiva’s tracking them as “Ben’s startup and Jenny’s startup.” Given that a Palantir spokesperson last year noted that the 22-year-old company had approximately 4,000 former employees, that means roughly 10% of all former Palantirians have gone on to launch a startup.

---

How to hire like Palantir

“Palantir was not a company I sought out,” says Matt Lynch. In 2014, he was a civilian engineer for the Navy, building software, when he decided it might be time for a change. “I’m just going to work at Google,” he thought. “Google seems like a great place to come up as a software engineer. It’s colorful and friendly and they have free food.” 

During the process, Palantir also reached out, intrigued by his experience building for the government. “I was like, ‘I don’t know what your company is, but I’m going to be in New York, so I’ll come by and check it out,’” he told them. During his Google interview, he says, “it was effectively the meat grinder where they make you sweat for hours, you’re grinding on the whiteboard without the interviewers really interacting with you at all.” 

The next day, he visited Palantir. “I remember those interactions much more distinctly than I do the Google ones,” he says, “because they were so much more personalized.”

Lynch didn’t get an offer from Google but he did from Palantir, where he stayed until 2021 when he left to cofound Sage, a hardware and software platform to deliver better care in assisted living facilities. 

Palantir’s hiring process hinges on providing candidates a real sense of the culture. “We had to get [recruits] into an office,” says Ross Fubini, founder and managing partner of the venture firm XYZ and a Palantir adviser since 2010. “They’d come in and see, ‘Oh, this is a jokey, joyful technology company, not a boring consulting one.’ They’d also see the intensity of the people.”

No matter how long it’d been since the Palantir alums I interviewed had gone through their interview process, they recalled key details. There were typically five interviews. Everyone is assessed on their technical acumen, but it wasn’t dispositive. “I remember basically getting punched in the face for 45 minutes,” says Manifest’s Frankel. 

“I don’t think I would be hired by Palantir today or any time in the last 10 years,” says Zach Romanow, thinking about how he didn’t have the technical background that has often been required to get hired at Palantir in the last decade. He spent 11 years at the company, from 2012 to 2023, before cofounding Fourth Age, a startup that offers Palantir customers specialized forward-deployed engineering to build complex applications on top of Palantir’s platforms. Romanow describes Palantir’s hiring ethos at the time as “Let’s try and hire smart, hardworking people with the right motivations and we will figure out how they can be useful.”

Multiple alums share their memories of what’s known as a “decomp interview.” Sage’s Lynch says, “You’d spend the hour talking through how you would effectively design an anti-money laundering system or something like that.” Others say their challenge was to optimize the operations of an elevator bank in an office building to minimize the wait during peak usage times or to design a subway system where a seat would always be ready for you. 

“The whole idea is can you think abstractly about a problem and can you take something that sounds impossible on its face and start to add structure and rigor to it so that you can turn it into bite-sized chunks that you could actually then go execute or test hypotheses,” Lynch says. “I had never been put in a position to do an interview like that before.” 

There’s also a “behavioral portion” of the interview. “They asked me pretty intense questions,” says August Sun Chen, who applied to work at Palantir in 2021 after graduating from Harvard University and spending a year in consulting. He’s now CEO of Hazel, an AI solution for government procurement. “They asked me to tell them the three decisions that made me who I am today.” 

All of it is designed to find people who have the grit to handle the challenges inherent in the way the company works. “One of the things we took from Palantir is insisting on a really rigorous intentionality in the interview process,” says John Doyle, who spent nine years at Palantir before cofounding Cape, a privacy-first wireless service. “Each interviewer knows which facet of the candidate they’re testing, and they use the same questions to test that facet over time so they can develop a little bit of internal data about what good looks like, the range of outcomes, and also how people perform over time based on how they did on various facets.”

As he notes, it’s a lot of work up front, but “you can have a pretty high success rate.”

“Figure it out”: Why Palantir mints entrepreneurial talent

“You spend the first week just working on a demo project as part of your onboarding,” says Alex Shieh, who dropped out of Brown University in 2025; he had applied to spend a semester at Palantir (one of the company’s many unconventional programs to bring in and assess talent) but was hired full time. “Then you get thrown into a deployment. The idea is that you’ll get the hang of it.” 

Palantir’s forward-deployed model for its implementation teams is the defining element of its corporate culture and also the reason it’s produced an outsize number of company founders. Within the implementation teams, there are two roles: forward-deployed engineers (FDEs), who are the most technically adept; and deployment strategists, who are more the account lead. Or as they’re known internally, deltas and echos. “We liked military technology,” says Chapter’s Blumenfeld-Gantz.

“Internally we kept inventing [titles],” admits XYZ’s Fubini. Take lawyers. “We don’t call them lawyers,” he says. “They’re ‘legal ninjas.’ Your job is not just to do the contract review”—and potentially slow things down. That’s what a lawyer might do. Legal ninjas ask, “How can we help you move forward? How can we decode the contract to be successful and win?”

Whatever the title, they all have the same goal: Figure out the customer’s problem and do whatever they have to do to solve it. “Everyone who worked at Palantir at a certain phase of the company has a story about a customer needing something and it was pretty wacky, but you just did it,” says Jason Hoch, a cofounder of Nominal, which helps hardware engineering teams, people who build such things as nuclear fusion reactors and satellites, test and deliver complex systems faster. While at Palantir, Hoch worked as a product developer, forward-deployed engineer, and a product development lead.

“One of the things I really loved about Palantir’s engineering culture, its whole company culture, was its bias toward action,” says Pablo Sarmiento, CEO of Avandar Labs, which makes software for social enterprises and nonprofits to manage their data and who worked on Palantir’s philanthropy team as an FDE from 2013 to 2017. “If something is wrong or missing, just fix it.”

Employees had been screened, of course, to

• be mission-driven;
• be intrinsically motivated, or in current tech argot, have “high agency“;
• have what Manifest’s Frankel calls “low ego, high ops tempo,” meaning they’re focused on delivering good outcomes for customers;
• be “learning machines”;
• solve those “decomp” exercises, which are a simulacrum of what you do on deployments into government departments and commercial enterprises.

So yes, even new employees, often in their twenties and sometimes in their first real job, are “given an insane amount of ownership” when they start, as Sage’s Lynch describes it, “and very little problem definition.” 

But when you hire people with these traits, “it would be dumb to micromanage them,” he adds, though at Sage he does seek to play a role in thinking through a problem before letting employees take ownership. 

Not that there’s no support structure for these deployments. “There was a really strong culture of collaboration and sharing via the workplace chat systems, lots of public channels where people can freely share problems and challenges without fear of looking bad,” says James Ding, CEO of Draftwise, which makes AI software for law firms and in-house legal teams to automate contract drafting, review, and negotiation. Despite Palantir’s secretive image, within the organization people are cc’d on hundreds of emails a day to stay in the loop, and Ding says he could email the New York office “distro list” to ask a question and get help as needed.  

But on balance, the job is, in sum, to figure it out, even if one worked in product development or a back-office department like finance. “They thought that I could come in and help do really unstructured things,” says Sage CEO Raj Mehra, who joined the finance team in 2013, after founding a healthcare startup that failed. “To be honest with you, that’s all I did at Palantir. I did the most unstructured projects, things that no one thought we needed to do and they just threw me in it.”

If you found that exciting, then you would do well. If you wanted someone to check your work or give you a requirements document that you just executed, Palantir “was going to be a really frustrating experience for you,” says Lynch, Mehra’s cofounder and Sage’s CTO.

Or as Hazel’s Chen says, “You either work three months and quit or you’re there for years.”

Extreme agency and its limits 

Let’s be frank: In too much of corporate America, This. Does. Not. Happen.

From Hollywood to consulting to finance, businesses are ever more risk averse to empower early-career professionals to take the reins on a big project. Their structures thwart it. Even in Big Tech, a young employee is going to be told what to build and how to do so. “You don’t have a ton of context about why,” says Andy Chen, who had a number of experiences before working for Palantir from 2015 to 2020. “You don’t get to exercise your creativity as much.” (He’s now CTO of Nira Energy, which helps clean energy developers, data centers, and utilities understand where there’s available capacity on the electric grid for new projects.)

These kinds of circumscribed career paths can be limiting. “When I was in consulting,” Manifest’s Frankel says, “somebody who was really trying to be a mentor but who scared the living daylights out of me instead, told me that you start in Excel and you spend the first two years of your career in Excel. If you get good enough at Excel, we move you to PowerPoint. And if you do PowerPoint well enough, you move over to writing proposals in Word. Then if you’re really successful in Word, eventually you move into Outlook. And Outlook is where you manage the client relationships. 

“He was saying this in all sincerity,” Frankel continues. “You can track your development and maturity as a consultant through the Microsoft Office suite of products. There was nothing more corrosive to me than that idea that I was going to chart my life [that way].”

No worry of anything so stultifying at Palantir. Jack Fischer, CTO of the agentic AI startup Credal and who worked at Palantir from 2017 to 2022, recalls one assignment “in a skiff”—in Washington, D.C.—“and everything is on fire all the time and nothing’s working, and it is a continuous, multilayered emergency that just needs nonstop creative problem solving.” Those kinds of assignments, says Sage’s Lynch, were “addicting, That’s how I ended up with so much [personal] growth, because I got addicted to that dopamine hit every six months of ‘Oh, here’s a new thing.’”

For most of Palantir’s history, the reason the company looked like “technology-powered consulting,” as XYZ’s Fubini describes it, is because the products were not truly ready for customers. That left it to those forward-deployed engineers and deployment strategists to do the requisite “duct taping” of the product, says Fourth Age’s Romanow.

That process of taking software that’s “amorphous clay” that needs to be shaped for a customer, as Draftwise’s Ding describes it, does two things:

It creates the opportunity for what Palantir calls repeatability, where one of those custom solutions can be offered to other companies. Ding says that he was the primary developer on a product for a banking client that was resold to seven more banks. These kinds of opportunities are how Palantir has been able to accelerate its growth and profitability in the way it has, increasing revenue 77% year over year and 20% sequentially in Q3 2025. GAAP net income in the quarter hit $476 million, a 40% margin. 

It’s also created a lot of future founders. 

The lessons they carried

In the past year, as Palantir has been on a heater, corporate America has decided it needs to hire its own forward-deployed engineers, with the Financial Times reporting that the FDE role had become the most popular new job title in business. 

This is, as you may guess, missing the point. “The forward-deployed engineers, which people talk about, people still get that wrong,” says XYZ’s Fubini. If “the DNA of the company is already set,” he notes, then the FDEs become “basically technical support people. No, they’re your core engineers and they’re on-site with customers, they’re bringing cupcakes into the break room. They’re there to get access to the problem and bring that knowledge back.”

Companies want the cachet (and market cap) they associate with Palantir’s FDE model, but almost every fiber of their being fights it. “Even in Silicon Valley,” says Nominal’s Hoch, “there’s this agency that investors give to founders to build something. But very quickly, companies—it’s almost the state of nature—put up a lot of boundaries and rules and processes.” He adds, “Falling into the ruts of corporate organizational norms makes it hard to achieve 10x better outcomes.”

What counts are the underlying principles, not titles, and Palantir doesn’t have a monopoly on them. 

“One thing Palantir did really well was what we called ‘seeking truth,’” says Chapter’s Blumenfeld-Gantz. “Are people actually using your product? What do they think? Are you actually talking to customers and embedded with customers? It’s shocking that most companies don’t actually do this well. They don’t get real feedback.” Or, he says, they talk only to people who have had good experiences. “They don’t ask, ‘What sucks about my product? What can I do better?’ [Most companies] are more focused on selling their product to customers than using customers’ experiences to improve their product. There’s a fear that by continually pressing customers for critical feedback you are highlighting negatives and jeopardizing revenue.

“What Palantir understood,” he adds, “is that no product is perfect, sophisticated buyers generally prefer honesty to bullshit, and the best long-term strategy for customer retention and product development is learning the actual truth about your product and continually improving it. The first question we always asked was, ‘What do you hate about this?’ That is a very valuable lesson, and it goes to having a thick skin. You have to be able to handle that.”

This probing curiosity to unlock valuable insights also builds confidence, which definitely helps when pursuing a startup. As does cultivating diverse points of view and a willingness to express your opinion. Alex Shieh left Palantir last August after just a few months to cofound the Antifraud Company, which roots out corporations cheating the government using both AI and investigative journalism. Shieh recalls that during his onboarding, “[CEO Alex] Karp said he likes to hire conservatives out of the Ivy League [because] they must be independent thinkers [overcoming] strong pressure to conform. … It’s also not the case that everybody at Palantir, or even most people at Palantir, are conservative or something. That’s a misconception.” 

Avandar’s Sarmiento says, “Whether you agree or disagree with Palantir, the one thing I always respected about them was that, at least during my time—I really don’t know about right now—they were very adamant and open about if you disagree with anything, you can speak about it. Palantir always encouraged that ‘question what you are doing’ approach. I grew personally and professionally so much thanks to Palantir that, in a sense, Palantir grew me into the kind of person who would no longer want to work at Palantir.”

What could derail Palantir’s founder factory?

Ross Fubini founded his venture firm XYZ in 2017 on the thesis that he would invest in Palantir alums who wanted to start companies because of everything he’d seen as a Palantir adviser. The whole point was that he anticipated something emerging that no one else had, thanks to his advisory role. “You could just see it through these very entrepreneurial, very technical groups, and that was how the whole organization was structured,” he says, adding, “I was talking about [Palantir] as an authoritarian democracy. There’s no question that Karp, then and now, is in charge of this business.”

Karp sets a clear strategic direction for the company. But the democracy is “leaders and teams within the organization self-organizing around problems and opportunities,” Fubini says. “There’s something here, go find it.”

If Palantir can maintain the equilibrium of that seeming oxymoron (authoritarian democracy), then its power will only grow, as will the network of employees turned founders.

It’d be far too early to suggest that anything threatens this hegemony now, but there are a handful of potential challenges worth watching closely.

The Palantir-diaspora relationship

Last year, Palantir sued two founding teams of its alumni alleging the theft of trade secrets and poaching employees. One source tells me that the company has sent other cease-and-desist letters to founders and reminded former employees of any non-solicitation agreements they signed.

Fubini sees both sides of the issue, given his dual roles. “I’ve been intimate through multiple of the cease-and-desist events,” he says. “To every one I’ve seen, Palantir believed it was ethically correct.”

As yet, neither Fubini nor Luba Lesiva, who backs only Palantir-led companies at Palumni VC, say they’ve seen any slowdown in startup formation that would signal a chilling effect. “We haven’t seen too much of the stepping on toes,” Lesiva says, noting that she has not encountered anyone actively looking to compete with Palantir, “but we’re also really open with our companies being like, ‘Hey, don’t be a fool’ when it comes to making sure they’re abiding by their exit documents and seeking legal advice if necessary.”

The current climate, though, does cast some shade on the sentiment many founders shared, best reflected by Angle Health cofounder Anirban Gangopadhyay: “They would always say that the only reason you should leave Palantir was to start your own company.”

Foundry eats the world

Thanks, then, to the product leverage Palantir has with Foundry and its AI platform (AIP), “in theory, everything could be a Palantir project,” Fubini acknowledges.

Lesiva says that of the companies she’s invested in and that she’s tracking, she sees three opportunities.

• Pick a niche: “Some founders choose to start small, [with] the specific part of a specific problem,” she says, citing Tamarack, a company building mill management software for logging companies to generate more revenue. “That’s just not going to be a giant focus of Palantir.”
• Do the opposite: Lesiva reminds me that “Palantir doesn’t collect data; Palantir handles the data that its customers already have.” That creates space in data collection or what Lesiva calls “the actioning of the data on the other side.”
• Partner: The company has programs for earlier-stage startups, including FedStart, to facilitate meeting government compliance standards, and Foundry for Builders, which forges a technology partnership. Blumenfeld-Gantz’s Chapter helped start the latter program and was its first customer. Palantir alum-founded Hence and Adyton (see dataset) were also among the first startups to participate in Foundry for Builders.

What now constitutes a startup that won’t compete with Palantir yet is something big enough to be worth doing? The answer could affect who leaves Palantir (and why) as well as who stays.

Keeping the talent pipeline wide open

Fubini ponders exactly what it was he saw in the mid-2010s bubbling up at Palantir and why those teams within the company were so successful. “One is just a relentless focus on talent,” he says. “Back then, it was really just we’re hiring young people, highly tactical out of universities, and that was the goal: Be a premier place. And a lot of attention went to doing this, building that recruiting group.” But, he adds, “I have to tell you, we were shit at—we’re still largely shit at—bringing in senior people because of this structure.”

Fubini emphasizes that what’s most needed is Palantir continuing to bring in earlier career and elite technical talent.

As Karp has publicly expressed dismay at campus activism in the past few years, Palantir has opened up new vistas, such as the semester at Palantir program; the Meritocracy Fellowship it’s introduced for high schoolers; the Neurodivergent Fellowship, which it debuted after Karp’s restless appearance at last December’s New York Times Dealbook Summit; and, most recently, the American Tech Fellowship for Veterans. Whether these programs can augment, much less replace, its previous campus recruiting efforts remains an open question.

I ask Fubini what he believes could slow Palantir down. “Fundamentally,” he says, “people just need to come there for the mission. Still. That they believe in something is a very ego-rich thing. We’re better, smarter, we’re going to solve these problems. And some of that’s the Americana and some is ‘I’m going to solve a really hard problem in oil and gas or insurance that nobody else is doing.’ The second part is [Palantir’s] got to stay on the technology edge. It’s got to be a place working on the hardest, best, most interesting problems, with the platform or otherwise.

“If you’ve got one or the other, you’re okay,” Fubini says. “But the perfect intersection is deeply technical and mission driven. Those are the people you want.”

View the full article