 |
|
|
|||||
Last week, I published a deep exploration into Palantir and its founder factory and how the companys power and success can be explained by its ability to attract elite talent and how it empowers them to develop their skills and learn new ones in the projects they pursue. That talent then goes on to found their own startups, invariably seeking to address hard, intractable problems much as they did in their work at Palantir. (In the few days since I published my first story, Ive found another 21 former Palantir employees turned founders, bringing what was already the largest public dataset of these people to 335. If you havent already, check it out here.) There are a number of high-profile companies founded by Palantir alums that many people have heard of. These include: Anduril, the defense contractor ($30.5 billion valuation), cofounded by Brian Schimpf, Matt Grimm, and Trae Stephens; Kalshi, the predictions market ($11 billion valuation), cofounded by Tarek Mansour; Eleven Labs, the voice AI platform ($6.6 billion valuation), cofounded by Mati Staniszewski; Handshake, the marketplace for early-career workers, colleges, and employers which has recently focused more on matching specialized talent with AI training opportunities ($3.3 billion valuation), cofounded by Garrett Lord; and Partiful, the planning tool for IRL experiences ($400 million valuation), cofounded by Shreya Murthy and Joy Tao. But there are so many fascinating stories among the cadre of startups founded and led by Palantir alums. The nine companies showcased below, which span healthcare, government services, cybersecurity, law, clean energy, hardware development, and not-for-profits, exemplify the power of being acculturated to finding a big, hard problemand having the skills to tackle it. Angle Health Palantir alum founders: Ty Wang (CEO), Anirban Gangopadhyay Other founders: None What Angle does: AI-native healthcare benefits platform, particularly serving employees of small and medium-size businesses Employees: Approximately 100 Funding: $197 million raised to date, including a $134 million Series B in December 2025 led by Portage, with funding also from Blumberg Capital, Y Combinator, and others Secret sauce: Creating a full-stack solution. To work toward achieving the companys goal of changing how people approach and access healthcare, as Wang says, and democratize access to the kinds of modern healthcare services . . . that are still not available to a lot of the people that really need them, Angle had to rebuild the technology infrastructure that powers the way that the vast majority of Americans access healthcare today, which is through their health plan. Angle has centralized data assets and focused on enabling AI-driven, human-in-the-loop workflows across its products and operations. That allows it to offer such things as digital behavioral health programs or digital pharmacies, for examplenewer services that have become more routine for large employers to include in their health plansand that can reduce the overall cost of care for the thousands of small businesses that Angle serves. One key learning from Palantir: Gangopadhyay explains that Angles culture encourages a lot of slow thinking and having discourse on a monthly basis to develop a clear plan, and then we’re very intentionally head-in-the-ground and hands-to-keyboards executing. Although Palantir itself is not structured this way in terms of a “monthly sync,” he adds that we’re very light on meetings. We leave the individuals to execute in their own way. That is spiritually aligned with how Palantir operates. Avandar Labs Palantir alum founders: Pablo Sarmiento (CEO and CTO) Other founders: None What Avandar does: Software for social enterprises and nonprofits to manage their data Employees: One Funding: Bootstrapped. In fact, Sarmiento says he will not raise equity-based funding, choosing instead to pursue non-dilutive capital sources, including revenue-based financing to align better with the goals of socially focused companies. Secret sauce: Think about it, says Sarmiento, we shouldn’t be building the software we need to fight a crisis during the crisis, referring to COVID as well as the work he did after he left Palantir, at Zenysis Technologies, helping create software so that the National Health Institute in Mozambique could successfully fight a cholera epidemic. It should exist. Avandar Labs lets not-for-profits and social enterprises build a unified data platform to integrate and analyze an organization’s program data. (It’s currently in beta but when complete, Sarmiento says it’ll be “customizable to any mission.”) The platform’s core technological difference is that it was built from the start for social sector use cases, such as “ensuring it can support epidemic response, humanitarian emergencies, and cross-sector coordination.” He promises it’ll be far cheaper than any alternatives, too. One key learning from Palantir: Bias towards action. Chapter Palantir alum founders: Cobi Blumenfeld-Gantz Other founders: Corey Metzman, former Presidential candidate (and current Ohio gubernatorial candidate), Vivek Ramaswamy What Chapter does: AI to help American seniors find the optimal Medicare plan at the lowest cost Employees: Approximately 200 Funding: $186 million raised to date, most recently at a valuation of approximately $1.5 billion. Investors include Stripes, XYZ, and Susa Ventures, among others. Secret sauce: Using AI to help seniors navigate Medicare. Chapters recommendation engine identifies which of the 24,000 Medicare options that exist is right for an individual customer, taking into account their doctors, prescriptions regimen, usual pharmacies, the benefits that are most important to them, their ability and willingness to pay, and more. Each one of those inputs is a huge data problem in and of its own, Blumenfeld-Gantz notes. It has an app that can determine from a picture of a users Medicare card which plan theyre on and then curate every single item that’s eligible for your plan and check them out without a litany of phone calls, he adds. Speaking of calls, Chapter ingests every phone communication its brokers have to assess if they’re making high-quality, compliant recommendations and offer real-time feedback. One keylearning from Palantir: Being relentless. Working in a regulated space, where you have to get federal and state licenses and get licensed by insurance carriers in every state you operate, its just not accepting no, says Blumenfeld-Gantz. [You have to be] really annoying to state departments of insurance until they take your call and move your paperwork forward. The way I think about it is that you have to make it less work for them to do what you want them to do. Status quo, the easier thing is for them to do nothing. So you have to change the status quo so its easier for them to do something than nothing. Draftwise Palantir alum founders: James Ding (CEO), Emre Ozen Other founders: Ozan Yalti (former senior associate at the global law firm Clifford Chance) What Draftwise does: AI software for law firms and in-house legal teams to automate contract drafting, review, and negotiation Employees: Approximately 60 Funding: $28 million raised to date, from Index Ventures, Y Combinator, and others Secret sauce: Every other well-funded legal tech company in the space is building an application layer tool trying to put LLMs inside of bespoke interfaces to try to increase productivity for lawyers, says Ding. Draftwise is a data platform. We started by recognizing that the pain point we wanted to solve was one where the challenge is that big-ticket deals require data, and if you can’t have the data, you can’t make good decisions. We started from that foundation, integrating data across a variety of silos, bringing it together, and shaping it into an ontology. Then we also happen to have interfaces to serve that data to people inside their workflow. For example, Ding cites an add-in for Microsoft Word that Draftwise made. You’re drafting a contract, you’re negotiating financial covenants, Draftwise can pull together into a single view all the data you need to actually make the decision of what covenants to give. One key learning from Palantir: The thing I wanted to bring was immense agency, immense accountability, a sense of high integrity, says Ding, but also high effort where we’re just getting things done, we’re doing it right, and we’re doing the best we can. Fourth Age Palantir alum founders: Zach Romanow, plus founding partners Jesse Rickard, Pete Mills, and Samuel Tarng Other founders: None What Fourth Age does: Specialized forward-deployed engineering for Palantir customers to build complex applications on top of Palantirs platforms Employees: More than 50 Funding: Bootstrapped Secret sauce: My first customer is really the engineers, says Romanow, the best and brightest FDEs, or the people that could become the best and brightest FDEs if they’re in the right place and have the right teams around them. . . . hire the best possible people that at scale provide differentiated outcomes for customers, and the customers will pay you accordingly. One key learning from Palantir: If you have a very, very high bar for the people . . . then A players want to join the A team, Romanow says. Let’s really stay true to our principles of what we know great looks like. Manifest Palantir alum founders: Daniel Bardenstein (CEO); Marc Frankel (former CEO) Other founders: N/A What Manifest does: Software and AI bill of materials to protect everything from healthcare systems to military aircraft Employees: Approximately 30 Funding: $21 million raised to date from such investors as AE Industrial Partners (Boeing’s venture arm), Palumni VC, XYZ, and others Secret sauce: Provides both vendors and buyers with visibility into the provenance of the elements in the software and AI they depend on to eliminate the risk of introducing a potentially calamitous vulnerability. Software is the only thing that we buy that you don’t get to know what’s in it, says Frankel. Everything else in our lives comes with an ingredients list. One key learning from Palantir: Low ego, high ops tempo.” Nira Energy Palantir alum founders: Andy Chen (CTO) Other founders: Chris Ariante (CEO, ex-Exxon Mobil), Andrew Martin What Nira does: Software for clean energy developers, data centers, and utilities that helps them understand where theres available capacity on the electric grid for new projects Employees: Approximately 30 Funding: $65.5 million from Energize Capital, Y Combinator, and others Secret sauce: Focusing on one of the most painful roadblocks to building renewables, the hidden pain point impeding the goal to accelerate America’s power grid to be fossil free as quickly as possible, as Chen says. That is what’s known in the energy business as interconnectionadding renewable projects to the grid. Nira’s built mapping tools to help developers identify sites with capacity and another one to estimate costs while a project is in queue to come online. One key learning from Palantir: Learning about transmission planning is a critical part to being successful at Nira, Chen says. If you’re not interested, you’re not going to be able to learn it. One thing that’s similar culturally between Palantir and the people we have here is this fundamental curiosity and willingness to learn about totally random stuff that will never help you in a future job, but you want to do it because you’re fundamentally interested in it. Chen adds that he’s now hiring for a forward deployed engineering role. Nominal Palantir alum founders: Jason Hoch Other founders: Cameron McCord (CEO, former Naval submarine officer, ex-Anduril), Bryce Strauss (ex-Lockheed Martin) What Nominal does: Software to help hardware engineering teams, people who build such things as nuclear fusion reactors and satellites, test and deliver complex systems faster Employees: Approximately 100 Funding: $102.5 million raised to date, from Sequoia Capital, Lightspeed, Lux Capital, Founders Fund, and others Secret sauce: Speed and solving the data challenges that hardware manufacturers face. When mechanical and electrical engineers work on hard hardware problems, they [also] have software problems, they have data infrastructure problems, Hoch says. We’re speeding up the workflows. We’re increasing the maximum complexity of what the hardware engineers and or customers can accomplish. When they finish a task or a simulation, they don’t need to crack open Claude Code to start understanding their data. It’s just right there in front of them. They’re able to ask the hard physics and engineering questions of the data. That’s the speed. For that data issue, when you’re building complex software systems, Hoch explains, you have this incredible toolkit of SaaS companies that have been building ways to make your job better for 30 years. Hardware engineers, by contrast, You’ll have 10,000 data points a second, a million data points a second coming off of a sensor, meaning that the nature of helping them process that data is not a solved problem. One key learning from Palantir: Remaining customer obsessed, remaining technically obsessed. Our customers are wildly technical. The things that I would have to teach people 12 years ago when I was onsite with a customer, these people already know it. It’s keeping us honest to making sure we’re really staying at the cutting edge. Sage Palantir alum founders: Raj Mehra (CEO), Matt Lynch (CTO) Other founders: Ellen Johnston (chief product officer) What Sage does: A hardware and software platform to deliver better eldercare, particularly in assisted living facilities Employees: More than 100 Funding: $59 million raised to date, from IVP, Friends & Family Capital, Maveron, and others Secret sauce: Building hardware to collect the critical data to support its software. How do we give caregivers better tools to care for residents? How do we give residents of these communities tools to call for help and get help when they need it? asks Mehra. Realizing that existing systems werent measuring relevant data, Sage has built Core, which tracks nurse calls and helps caregivers manage tasks, which operators can then “use to improve quality of care and caregiver performance,” Lynch notes. It also built Detect, which is AI-powered fall detection that enables care providers to respond proactively to those kind of emergency events. We can measure and pull in all of the telemetry from the physical devices that we’re deploying, Mehra adds. Based on all of that, you can then synthesize it and provide value to folks up the value chain. One key learning from Palantir: One of us is responsible for every single person we bring in, says Mehra. One of Palantirs founders interviewed every hire for a long time. We haven’t departed from that, he continues, and I don’t think we ever should because it’s how we keep the culture intact. Adds Lynch: Then, if we make a mistake, we own it . . . when bets don’t pay off, we can’t sacrifice the culture for that.
Category:
E-Commerce
A new book by a former Deloitte executive turned workplace well-being expert argues exactly that In her new book Hope Is the Strategy, Jen Fisher, an expert on workplace well-being and human sustainability, makes a clear and timely case that hope isn’t a soft skill or a leadership afterthought; it’s a practical, learnable approach to navigating uncertainty and building healthier, more resilient organizations. In the following excerpt, Fisher draws on her personal experience grappling with burnout, as well as her research on well-being, leadership, and corporate culture, to reframe hope as something we can all learn and implement for ourselves and those we work with. We’ve long misunderstood hope in the workplace. We’ve treated it as wishful thinkinga nice-to-have feeling that emerges when things are going well. But research from psychologist C.R. Snyder reveals something far more powerful: Hope is a cognitive process with three essential components: goals (what we want to achieve), pathways (our ability to identify routes to those goals), and agency (our belief that we can pursue those paths). This isn’t passive optimism; it’s an active strategy for navigating uncertainty and driving meaningful change. After my own experience with burnout, I discovered that hope isn’t what you turn to after strength failshope is the strength we’ve been looking for all along. It’s not the light at the end of the tunnel; it’s the torch we need to lead others through it. And when organizations embed hope into their leadership practices and culture, they unlock something remarkable: the capacity to transform not just how people feel about work, but what they can actually accomplish together. As more organizations prioritize helping their employees become healthier, more skilled for the future, and connected to a sense of purpose and belonging, they have an opportunity to instill hope in leadership and encourage it in workers. A roadmap for the future A leader who has hope can map out a path for an employee, offering a solid roadmap rather than an empty promise. They might say, “I can’t promise you complete job security, but I can provide you with the skills that will make you attractive in the job market.” That, in turn, helps foster hope in the worker, because they know that they’ll have more tools in their success toolkit, no matter what the future holds. That’s not just a win for the individual, but for the group. An organization (of any typeit could also be a community, or a family) filled with people tapped into their meaning and purpose is stronger than one made up of disengaged, unhealthy, and unhappy people. In fact, hope is a strategy for a variety of prevalent workplace problems: It can improve mental well-being and stress management; it can drive action and reduce catastrophic thinking; and it can help overcome the disengagement crisis at work. What’s more, hope will support our transition to a more human-centered workplace as AI takes on the more mundane, tactical aspects of work. Creating new ripples from leadership on down is possibleand as with the negative ones, it starts with modeling behaviors to set the tone for your team and your peers. That is, modeling the sustainable work behaviors and values that will drive purpose and well-being. Here are four examples: 1. Get clear on what your own boundaries are If you’re following someone else’s vision of success instead of your own, you’re going to end up miserable and probably burned out. So take that PTOreally. The company will not crumble without you. And don’t answer that email at midnightreply in the morning, during work hours. A leader who actually sets healthy boundaries and lives by them gives employees permission to do the same. As I reevaluated the role that work played in my life, I set my own new boundaries. I got clear on what my definition of success was, instead of allowing the external world to define that for me. And I brought hope into my life: I started each day with a set of “what if” questions, looking at the day ahead through the lens of possibility: What if this goes right? What if I do things this way? Then I’d end each day with reflection: How did it go? It helped me to see challenges as an opportunity for change. Here are some other daily practices I put in place, all of which I still follow today: Treat sleep as a nonnegotiable. I protect my eight hours like the business asset it actually is, recognizing that sleep isn’t a luxury but the foundation that makes everything else possible. Schedule humanity into the calendar. Not vague “personal time” but specific blocks for connections that make me human: dinner with my husband, phone calls with friends, reading fiction that has nothing to do with work. Incorporate daily recovery rituals. Three-minute breathing breaks between meetings, a proper lunch away from my desk, a brief walk outside to reset my nervous systemthese small moments of renewal prevent depletion from accumulating. Defend the calendar against the tyranny of urgency. Breakfast, lunch, dinner, exercise, and sleep aren’t just activities to fit around “real work”they comprise the immovable infrastructure that sustains my performance. Everything else has to work around them, not the other way around. 2. Embrace the unknown When we temporarily suspend our need for certainty, a different kind of productivity emerges. I call these my Possibility Days: Once a week, I grant myself permission to coexist with uncertainty. Instead of trying to control outcomes, I deliberately seek experiences with unknown results. I have conversations without preparing talking points. I explore ideas that seem impractical. I follow curiosity down rabbit holes without worrying where they lead. My most innovative solutions and deepest insights almost always trace back to these deliberate ventures into possibility thinking. 3. Walk the walk The old ways of leading through power and control are giving way to something more human, more hopeful, and more whole. The future of leadership isn’t just about what we doit’s about how we show up, how we hold space for both struggle and possibility, and how we cultivate well-being as a vital way of being. There’s this old thinking that we should check our feelings or emotions at work. It’s basically telling people: Don’t show up as who you truly are. When leaders normalize having no energy, no life, no nothing beyond work, it becomes not just accepted but expected. Emotions, whether they’re positive or negative, are really a sign of the things we care aboutand when we’re told not to bring emotions into the workplace, it stunts creativity, growth, innovation, connection, and understanding. The answer is simple: Show your emotions. Your employees look to you to set the pace, tone, and stakes of the team and the work being done. Be vulnerable and authentic about when you’ve made a mistake, when you said one thing and you did another, when you screwed up. Your actions show themthat decisions to support their own health and well-being and career growth aren’t going to be viewed negatively or make it seem like they’re less committed to their work. 4. Build teams grounded in trust True organizational and individual success depends on teams built on mutual trustteams that prioritize deep relationships alongside personal well-being. Trust-based teams require leaders who actively invite people to show up authentically and provide genuine support when they do. This means fostering psychological safety where team members feel confident giving honest feedback, taking calculated risks, learning from missteps, and growing from challenges rather than facing punishment for them. Organizations with the strongest well-being cultures maintain ongoing dialogue between leaders and team members. Within trust-based environments, people develop a growth-oriented perspective. Colleagues treat each other with genuine care and respect, creating workplaces rooted in kindness. This positive energy extends far beyond individual teams, helping organizations attract diverse talent, improve retention, spark innovation, and build lasting resilience.
Category:
E-Commerce
When one of the founders of modern AI walks away from one of the worlds most powerful tech companies to start something new, the industry should pay attention. Yann LeCuns departure from Meta after more than a decade shaping its AI research is not just another leadership change. It highlights a deep intellectual rift about the future of artificial intelligence: whether we should continue scaling large language models (LLMs) or pursue systems that understand the world, not merely echo it. Who Yann LeCun is, and why it matters LeCun is a French American computer scientist widely acknowledged as one of the Godfathers of AI. Alongside Geoffrey Hinton and Yoshua Bengio, he received the 2018 Association for Computing Machinerys A.M. Turing Award for foundational work in deep learning. He joined Meta (then Facebook) in 2013 to build its AI research organization, eventually known as FAIR (Facebook/META Artificial Intelligence Research), a lab that tried to advance foundational tools such as PyTorch and contributed to early versions of Llama. Over the years, LeCun became a global figure in AI research, frequently arguing that current generative models, powerful as they are, do not constitute true intelligence. What led him to leave Meta LeCuns decision to depart, confirmed in late 2025, was shaped by both strategic and philosophical differences with Metas evolving AI focus. In 2025, Meta reorganized its AI efforts under Meta Superintelligence Labs, a division emphasizing rapid product development and aggressive scaling of generative systems. This reorganization consolidated research, product, infrastructure, and LLM initiatives under leadership distinct from LeCuns traditional domain. Within this new structure, LeCun reported not to a pure research leader, but to a product and commercialization-oriented chain of command, a sign of shifting priorities. But more important than that, theres a deep philosophical divergence: LeCun has been increasingly vocal that LLMs, the backbone of generative AI, including Metas Llama models, are limited. They predict text patterns, but they do not reason or understand the physical world in a meaningful way. Contemporary LLMs excel at surface-level mimicry, but lack robust causal reasoning, planning, and grounding in sensory experience. As he has said and written, LeCun believes LLMs are useful, but they are not a path to human-level intelligence. This tension was compounded by strategic reorganizations inside Meta, including workforce changes, budget reallocations, and a cultural shift toward short-term product cycles at the expense of long-term exploratory research. The big idea behind his new company LeCuns new venture is centered on alternative AI architectures that prioritize grounded understanding over language mimicry. While details remain scarce, some elements have emerged: The company will develop AI systems capable of real-world perception and reasoning, not merely text prediction. It will focus on world models, AI that understands environments through vision, causal interaction, and simulation rather than only statistical patterns in text. LeCun has suggested the goal is systems that understand the physical world, have persistent memory, can reason, and can plan complex actions. In LeCuns own framing, this is not a minor variation on todays AI: Its a fundamentally different learning paradigm that could unlock genuine machine reasoning. Although Meta founders and other insiders have not released official fundraising figures, multiple reports indicate that LeCun is in early talks with investors and that the venture is attracting atention precisely because of his reputation and vision. Why this matters for the future of AI LeCuns break with Meta points to a larger debate unfolding across the AI industry. LLMs versus world models:LLMs have dominated public attention and corporate strategy because they are powerful, commercially viable, and increasingly useful. But there is growing recognition, echoed by researchers like LeCun, that understanding, planning, and physical reasoning will require architectures that go beyond text. Commercial urgency versus foundational science:Big Tech companies are understandably focused on shipping products and capturing market share. But foundational research, the kind that may not pay off for years, requires a different timeline and incentives structure. LeCuns exit underscores how those timelines can diverge. A new wave of AI innovation:If LeCuns new company succeeds in advancing world models at scale, it could reshape the AI landscape. We may see AI systems that not only generate text but also predict outcomes, make decisions in complex environments, and reason about cause and effect. This would have profound implications across industries, from robotics and autonomous systems to scientific research, climate modeling, and strategic decision-making. What it means for Meta and the industry Metas AI strategy increasingly looks short-term, shallow, and opportunistic, shaped less by a coherent research vision than by Mark Zuckerbergs highly personalistic leadership style. Just as the metaverse pivot burned tens of billions of dollars chasing a narrative before the technology or market was ready, Metas current AI push prioritizes speed, positioning, and headlines over deep, patient inquiry. In contrast, organizations like OpenAI, Google DeepMind, and Anthropic, whatever their flaws, remain anchored in long-horizon research agendas that treat foundational understanding as a prerequisite for durable advantage. Metas approach reflects a familiar pattern: abrupt strategic swings driven by executive conviction rather than epistemic rigor, where ambition substitutes for insight and scale is mistaken for progress. Yann LeCuns departure is less an anomaly than a predictable consequence of that model. But LeCuns departure is also a reminder that the AI field is not monolithic. Different visions of intelligence, whether generative language, embodied reasoning, or something in between, are competing for dominance. Corporations chasing short-term gains will always have a place in the ecosystem. But visionary research, the kind that might enable true understanding, may increasingly find its home in independent ventures, academic partnerships, and hybrid collaborations. A turning point in AI LeCuns decision to leave Meta and pursue his own vision is more than a career move. It is a signal: that the current generative AI paradigm, brilliant though it is, will not be the final word in artificial intelligence. For leaders in business and technology, the question is no longer whether AI will transform industries, its how it will evolve next. LeCuns new line of research is not unique: Other companies are following the same idea. And this idea might not just shape the future of AI researchit could define it.
Category:
E-Commerce
All news |
||||||||||||||||||
|
||||||||||||||||||