How Palantir built the ultimate founder factory | Nabeel S. Qureshi (founder, writer, ex-Palantir)
This episode features Nabil Qureshi, a founder, writer, researcher, and former Palantir forward-deployed engineer, discussing Palantir's unique culture, hiring practices, and the innovative forward-deployed engineer role. He shares insights on building products, talking to customers, and lessons for aspiring founders.
Deep Dive Analysis
20 Topic Outline
Nabeel Qureshi's Background and Palantir's Impact
Palantir's Unique Culture and Hiring Philosophy
Key Traits Palantir Seeks in Employees
Palantir's Approach to Titles and Career Ladders
The Forward-Deployed Engineer Role at Palantir
Principles Behind Palantir's Platform Success
Understanding Palantir's Gotham and Foundry Platforms
The Ontology Concept in Palantir's Data Platform
Life and Impact of a Forward-Deployed Engineer
Balancing Custom Solutions with Product Vision
Implementing a Forward-Deployed Engineering Model
Evolution of Forward-Deployed Engineers at Palantir
The Power of Data Ingestion, Cleaning, and Analysis
Hiring for Startups: Mission Alignment and Drive
Distinctive Traits of Palantir Product Managers
Navigating the Moral Questions of Working at Palantir
Key Advice for New Startup Founders
Useful AI Tools and Workflows
Contrarian View: The Value of College
Lightning Round: Books, Movies, and Life Motto
5 Key Concepts
Bat Signal Recruiting
This is a recruitment strategy where a company puts out a distinctive message or set of values that intentionally turns some people off, but strongly attracts the specific type of talent they truly desire and who are aligned with their mission. The key is that it must repel some to effectively attract others.
Murder Board
An army-derived process for new projects where a two-page plan is written, and three or four smart individuals unfamiliar with the project are invited to rigorously critique and tear apart the plan. This ensures thorough vetting and strengthens the project's vision, goals, and tactics.
Forward-Deployed Engineer (FDE)
An engineer who works directly on-site with a customer, often spending days in their building, to understand and solve their problems. FDEs are empowered to build new software solutions if needed, rather than just deploying existing products, fostering deep customer empathy and rapid iteration.
Product Leverage
A metric used by Palantir to measure success, essentially revenue per engineer. The goal is to maximize this number by building powerful, generalizable products that reduce the need for a large number of engineers per customer, indicating a shift from a services to a product business model.
Ontology (Palantir)
A core piece of Palantir's Foundry platform that involves taking disparate data tables (often with obscure names) and mapping them to human-understandable concepts like 'part,' 'work order,' or 'aircraft.' This creates a clear, hierarchical view of data, making it accessible and actionable for non-technical users without needing complex queries.
11 Questions Answered
Palantir achieves tactical outcomes for large customers, typically Fortune 50 companies and governments, through its data platform. It offers Gotham for intelligence and defense, and Foundry for commercial use cases.
Palantir screens for independent-minded individuals who aren't afraid to push back, possess broad intellectual interests, and are intensely competitive with a 'win-at-all-cost' mentality. They also sought talent outside the traditional tech ecosystem, like former military personnel.
Palantir avoids traditional titles to prevent unproductive internal competition, gaming the system, and optimizing for metrics rather than actual product improvement. This fosters a meritocratic environment where individuals must continuously earn their place and responsibilities.
An FDE is an engineer who works on-site with customers, often for days a week, to deeply understand their problems and build software solutions alongside them. This role ensures direct customer empathy, rapid feedback loops, and the development of highly relevant products.
Palantir achieved this by having incredible product development talent that took internal tools used by FDEs to create customer value and unified them into generalizable products like Foundry. This process was driven by a mandate to get customers using these internal tools and a focus on solving common data integration pain points in large organizations.
Both are data platforms with data ingestion, mapping, and UI layers. Gotham is optimized for military, defense, and intelligence use cases, emphasizing map-based workflows and graph-based analysis (e.g., finding connections in networks). Foundry is optimized for commercial use cases, focusing more on traditional data queries, tables, and analytics, without as much emphasis on graph analysis.
Palantir PMs are almost exclusively internal promotions who first proved themselves as forward-deployed engineers. This ensures they have deep customer empathy, a strong drive to get things done, and a practical understanding of how to build and deploy solutions, avoiding a purely theoretical approach to product management.
Startups should be willing to pivot to solve a massive burning problem for a customer, even if it deviates from their initial product vision, rather than being too stuck on their own ideas. It's crucial to test early by asking customers to pay a lot of money; if they say no, move on to a different problem.
Nabeel suggests acknowledging the upside (e.g., COVID response, cancer research) and engaging with the 'gray zones' rather than disengaging entirely. He believes it's important to be in the room to potentially improve processes and reduce harm, and that tech employees should actively consider the political and societal impact of their work.
It's crucial to hire people who genuinely care about the mission and are willing to go the extra 20% beyond just checking boxes. Aggressively filter for mission fit and ask about past experiences where they worked hardest to achieve something, as this reveals motivation and drive beyond just skills.
Nabeel believes going to college is highly valuable, especially for deep friendships, intellectual exploration, and self-discovery, which are hard to replicate later in life. He advises against dropping out unless there's a very compelling reason, despite common tech industry sentiment.
27 Actionable Insights
1. Validate with Customer Payments
When launching a new product, ask potential customers to pay a significant amount of money upfront. If they decline, quickly pivot to a different problem, as this indicates a lack of perceived value.
2. Pivot to Customer’s Big Problem
Be willing to abandon your initial product vision and focus entirely on solving a major, stated problem an enterprise customer has. This often uncovers a larger, more viable market opportunity than your original idea.
3. Hire for Extreme Drive
Prioritize candidates who demonstrate exceptional care and commitment, going the “extra 20 percent” to achieve outcomes. Screen for mission alignment and ask about their most challenging accomplishments and motivations.
4. Embrace Forward-Deployed Engineers
For large, complex customer engagements with high revenue potential, embed technical engineers directly at customer sites, working alongside them. This builds deep trust and allows for the creation of bespoke solutions that can later inform core product development.
5. Fast, In-Person Iteration
Implement rapid feedback loops by physically working with customers, building solutions, getting immediate feedback, and iterating multiple times per week. This hands-on approach fosters trust and accelerates product development significantly.
6. Productize Data Integration Pain
Recognize that most data work in large organizations is spent on access, cleaning, joining, and normalizing. Focus product efforts on streamlining these foundational data integration steps to make the entire process easier.
7. Cultivate Distinctive Internal Culture
Build a strong, unique internal culture with high trust among team members. This creates clear benchmarks for excellence and attracts talent deeply aligned with the company’s specific mission.
8. Promote PMs from Field
Only promote or hire Product Managers who have demonstrated their capabilities in customer-facing, problem-solving roles like forward-deployed engineers. This ensures strong customer empathy and a results-driven mindset, avoiding theoretical product management.
9. Challenge Product Principles
When defining project principles, ensure they are specific and potentially controversial enough that some people might disagree. Avoid generic statements that lack true guidance or differentiation.
10. Eliminate Competitive Titles
Consider removing traditional hierarchical titles to prevent internal competition and “gaming the system” for promotions. This fosters a meritocratic environment where individuals must continuously earn their place based on performance.
11. Deeply Understand Customer Business
Gain a profound understanding of your customers’ underlying business models and operational dynamics. This often reveals non-obvious problems and significant product opportunities.
12. Leverage AI for Cheaper FDEs
Recognize that AI tools are making it easier and more cost-effective for individuals to be highly technical and build features. This reduces the cost of implementing a forward-deployed engineering model, making it viable for more startups.
13. Engage Messy Real-World Problems
Focus startup efforts on solving complex, “messy” problems in traditional sectors like healthcare or government. The rise of LLMs and increased openness to tech now make these challenging areas more accessible for founders.
14. Aim for “Chartres” Excellence
Adopt the motto “aim for Chartres,” striving to create work that surpasses the very best achievements. This mindset encourages continuous pursuit of excellence rather than settling for adequate results.
15. Develop Deep User Empathy
Cultivate the ability to deeply understand and imagine the perspectives and motivations of your users and customers. This profound empathy is crucial for building truly effective and resonant products.
16. Use “Murder Boards” for Projects
For new projects, create a “murder board” by drafting a two-page plan and inviting smart, unbiased colleagues to rigorously critique your vision, goals, and tactics. This process helps identify weaknesses and refine strategies early.
17. Continuously Reassess AI Alignment
Regularly evaluate your career and work every few months to ensure alignment with the evolving landscape of AI. Focus on high-potential opportunities that leverage the increasing technological leverage provided by AI.
18. Become a “Cyborg” with AI
Actively integrate and adopt AI tools into your daily workflows to become a “hybrid cyborg.” Early and deep fusion with powerful AI technologies provides a significant advantage.
19. Utilize AI for Transcription
Use AI tools like Whisper Flow to transcribe spoken words into text, especially for lengthy prompts when interacting with LLMs. This can significantly speed up AI interaction and iteration.
20. Employ AI for Code/File Ops
Explore AI tools such as Claude Code that operate directly on your file system from the terminal, generating code and managing files. These “guided agents” streamline development workflows.
21. Leverage LLMs for Quick Scripting
Use LLMs for rapid scripting tasks, such as classifying transactions for taxes based on metadata. This demonstrates how AI can quickly automate mundane, data-heavy personal or work tasks.
22. Prioritize College for Growth
Value college as a unique period for forming deep friendships, intellectual exploration, and self-discovery, which is difficult to replicate later in life. Avoid dropping out without a truly compelling reason.
23. Network Actively
Proactively reach out to interesting individuals via email, introduce yourself, and say hello. Meeting new people and receiving emails can be a valuable source of energy and opportunities.
24. Read “Impro” for Social Dynamics
Read “Impro” by Keith Johnstone to gain insights into creativity, social behavior, and how to better understand and interact with people, consciously modulating your own presence.
25. Study “Henriette” for Power
Read Shakespeare’s “Henriette” history plays (Henry IV, V, VI) for profound insights into power, politics, and the sacrifices of leadership, relevant for understanding prominent figures and societal dynamics.
26. Read “High Output Management”
Read Andy Grove’s “High Output Management” in its entirety to understand his thought process and the origins of his management principles, beyond just summaries.
27. Maintain Mental Sanity
Regularly disengage from work and technology to “touch grass” and prioritize your mental well-being.
10 Key Quotes
30% of PMs that leave Palantir start a company.
Lenny
You need something that actually a lot of people are going to go like why are you why are you taking this principle? This seems wrong to me.
Nabeel Qureshi
The average kind of deal that Palance had was very large right it's in the many many millions of dollars which means that you could kind of pay for this as part of the thing that the customer got.
Nabeel Qureshi
The actual analysis is actually just the tip of the iceberg it's kind of the last five or ten percent and the 95 percent before that is I'm gaining access to the data I am cleaning the data I'm joining the data I'm normalizing it putting it all in the same format.
Nabeel Qureshi
Palantir's biggest competitor is a company rolling its own solution.
Nabeel Qureshi
The thing that is really hard to find is somebody who really really cares a lot about doing the thing and will go that kind of extra 20 percent.
Nabeel Qureshi
Disengagement isn't the answer.
Nabeel Qureshi
Your p successes goes up the more bets you make and it's sort of a function of how many bets you make and the probability of success goes into your bets.
Nabeel Qureshi
The cost of doing things like forward deployed engineering has fallen by maybe five to ten x now at least.
Nabeel Qureshi
The chess players who succeeded the most in the mid-2010s especially were the ones who were really early adopters of neural network based chess engines.
Nabeel Qureshi
3 Protocols
Project Murder Board Process
Nabeel Qureshi- Write up a two-page plan for the new project, detailing vision, goals, and tactics for the next three months, including principles.
- Invite three or four smart individuals who know nothing about the project.
- Their job is to rigorously critique and tear apart the plan, testing its limits and assumptions.
Forward-Deployed Engineer (FDE) Rapid Iteration Cycle
Nabeel Qureshi- Go into the customer's building on Monday for meetings and to understand problems.
- Build something on Monday night based on customer needs.
- Show the built solution to somebody on Tuesday and get feedback.
- Iterate on the solution on Tuesday night based on feedback.
- Show the updated solution to somebody on Wednesday and get feedback.
- Iterate on the solution on Wednesday night.
- Repeat this cycle, getting 4-5 feedback and iteration loops every week.
Key Principles for Implementing Forward-Deployed Engineers
Nabeel Qureshi- Ensure FDEs are real engineers capable of building new products themselves, not just deploying existing software or acting as solutions architects.
- Prioritize in-person presence to build close personal bonds and trust with customers, which is significantly more effective than remote interactions.
- Cultivate a deep understanding of the customer's business dynamics and problems, including counter-intuitive aspects, to identify true pain points and opportunities for software solutions.