Lenny's PodcastNabeel Qureshi: How Palantir's bat signal forged founders
Through forward-deployed engineers embedded at Airbus and the NIH; Palantir built Foundry and Gotham on data plumbing, breeding mission-fit founders.
CHAPTERS
- 0:57 – 6:14
Why Palantir is a “founder factory” (and what this episode will unpack)
Lenny frames the episode around a striking stat: a huge share of Palantir PMs go on to found companies or get promoted quickly. He introduces Nabeel’s background and sets up the core themes: Palantir’s operating model, hiring culture, and the forward-deployed engineer (FDE) role.
- •Palantir ranks #1 in PM-to-founder rate (30% of departing PMs start companies)
- •Nabeel’s background: FDE work across public health, COVID response, AI in drug discovery
- •Promise of an inside look at how Palantir hires, builds product, and scales
- •Episode roadmap: data platform, customer immersion, and founder training effects
- 6:14 – 7:32
What Palantir actually does: outcome-driven deployments on a data platform
Nabeel gives the simplest accurate description: Palantir tactically delivers customer outcomes, usually through a best-in-class data platform. He distinguishes the two flagship products—Gotham and Foundry—and explains Palantir’s focus on very large customers.
- •Palantir sells a data platform but positions work around outcomes, not infrastructure
- •Gotham: optimized for defense/intelligence use cases
- •Foundry: optimized for commercial/enterprise use cases
- •Enterprise focus: governments and Fortune-scale organizations
- 7:32 – 12:31
The people and culture: independent minds, intellectual range, and competitive intensity
The conversation turns to Palantir’s unusual culture and why it attracts a distinctive talent pool. Nabeel describes the traits Palantir screens for and how founder-led interviews reinforced a strong “bat signal.”
- •Hiring signals: independent-minded pushback, broad intellectual curiosity, high competitiveness
- •Founder interviews as a filtering mechanism and cultural amplifier
- •“Bat signal” recruiting: being polarizing is a feature, not a bug
- •Early mission framing: defense/intelligence focus and “save the Shire” ethos
- 12:31 – 16:12
Internal operating rigor: “murder boards” and principles people can disagree with
Nabeel explains a planning ritual used to stress-test projects: the murder board. The key idea is to surface flaws early and use explicit, controversial principles rather than bland ones like “move fast.”
- •Murder board: outsiders tear apart a 2-page plan to find weaknesses
- •Principles should be specific enough to invite disagreement
- •Fast iteration is enabled by aggressive critique and clarity
- •Early pressure-testing reduces wasted execution
- 16:12 – 19:11
Why Palantir minimized titles (and what it changed in incentives)
Palantir avoided formal titles to reduce promotion gaming and unproductive internal competition. Nabeel also notes trade-offs: status games can reappear as proximity to powerful execs, but fluid responsibility and merit-based leadership become easier.
- •No-titles philosophy: reduce Goodhart’s Law and promotion-driven behavior
- •Avoids incentives like “start a new product to get promoted” dynamics
- •Fluid leadership: easier to reassign responsibility when performance changes
- •Downside: status competition can shift to inner-circle politics
- 19:11 – 26:21
Forward-deployed engineers (FDEs): the core mechanism behind Palantir’s edge
Nabeel defines the forward-deployed engineer role: real engineers embedded on-site with customers multiple days per week. He distinguishes FDE types and explains why this model created exceptional customer empathy and execution speed.
- •Two engineering tracks: core product engineers vs. forward-deployed engineers in the field
- •FDEs sit alongside customers (desk, badge, access) to solve problems directly
- •Two FDE archetypes: deeply technical SWE vs. technical-adjacent, high-situational-savvy profiles
- •Customer immersion creates fast feedback loops and deep trust
- 26:21 – 30:00
From “services” to product: turning bespoke wins into platform capabilities
Lenny challenges the classic critique that Palantir is ‘consulting in disguise.’ Nabeel explains how Palantir productized internal tools (used by FDEs) into Foundry, driven by top-tier PD talent and enforced customer usage.
- •Misconception: “sparkling Accenture”; rebuttal includes product demos and margin profile
- •Internal tooling → customer tooling: ‘we are our own first customer’
- •Leadership mandate (e.g., requiring customers to use the tools) forced usability/reliability rigor
- •Palantir’s ‘secret’: deep understanding of data pain inside large organizations
- 30:00 – 38:38
Gotham vs. Foundry, explained simply + the Ontology breakthrough
Nabeel breaks the platforms into a practical mental model: ingest, map, and build user-facing workflows. He then tells the Airbus story that led to a major Foundry differentiator—Ontology—mapping alien tables into human concepts.
- •Platform pyramid: ingestion layer, mapping layer, UI/workflow layer
- •Gotham emphasizes maps and graph/network analysis; Foundry emphasizes table/SQL-style analytics
- •Airbus factory workflow: turning SAP tables into ‘aircraft/work orders/parts’ concepts
- •Ontology: human-legible object model becomes a core differentiator in Foundry
- 38:38 – 50:42
Life as an FDE: extreme travel, in-person trust, and rapid iteration cadence
Nabeel describes the operational reality: frequent travel, sometimes last-minute, because being in person massively accelerates trust and progress. He highlights how weekly build-demo-iterate loops compound into dramatic outcomes (e.g., Airbus ramp-up).
- •Aggressive travel culture pre-2020; COVID/IPO introduced more controls
- •In-person time builds trust faster than Zoom-based engagements
- •Iteration rhythm: build at night, demo next day, repeat multiple times per week
- •Embedded work can directly impact major operational metrics (e.g., production ramp)
- 50:42 – 53:16
How to apply the FDE model elsewhere + what changed at Palantir over time
Nabeel advises founders on when FDEs make sense and how to adapt the model for smaller deal sizes. He also explains Palantir’s evolving ‘product leverage’ goal: increasing revenue per engineer by reducing the engineers-per-customer ratio.
- •True FDE motion requires willingness to invest heavily; deal size must support it
- •Scaled-down version: 1 person supports multiple customers rather than deep embed
- •Key requirement: FDEs must be real builders empowered to create new solutions
- •Palantir trend: fewer engineers per customer as product leverage increased
- 53:16 – 59:24
The underappreciated moat: ingesting, cleaning, joining, and governing enterprise data
Nabeel explains why data work is mostly ‘below the waterline’: access, permissions, cleaning, normalization, and joining consume the majority of effort. He describes how Foundry productized these primitives and why organizational politics make it hard.
- •The iceberg: analysis is the last ~5–10%; the rest is access + prep + joining
- •Common pain: weeks to get data access; then data isn’t readily queryable or intelligible
- •Foundry primitives: universal connectors, previews, scheduled ingestion, easier joins for non-technical users
- •Real competitor is often ‘build it in-house’ + internal gatekeeping incentives
- 59:24 – 1:10:02
Hiring for the ‘extra 20%’ + why Palantir PMs are different
Nabeel shares hiring lessons for mission-driven startups: skills matter, but motivation and ownership are the real differentiators early on. He then explains Palantir’s PM funnel: PMs were typically proven FDEs, highly trusted by engineers, and allergic to process-only PMing.
- •Early team hiring focus: people who push for outcomes and go beyond ‘checkbox’ work
- •Mission alignment as a screening tool; look for personal reasons someone cares
- •Palantir PMs were often internal promotions from BD/FDE roles
- •PM success depended on deep customer empathy and strong trust with engineers
- 1:10:02 – 1:15:46
The moral question: defense work, disengagement vs. engagement, and tech’s political reality
Lenny raises the ethical controversies around Palantir. Nabeel offers a framework: many projects have clear positive impact, and for gray areas, disengagement is rarely the best default—responsible engagement can improve outcomes and reduce harm.
- •Positive-impact examples: COVID response, medical research, life-saving public health efforts
- •Gray-zone reality: defense tech can increase precision and reduce errors vs. older methods
- •Tech can’t avoid politics; society now forces accountability and scrutiny
- •Personal stance: actively think through trade-offs rather than reflexively opt out
- 1:15:46 – 1:37:28
Building a startup now: fast bets, strong culture, AI workflows + contrarian & lightning round
Nabeel distills operating principles he’s carrying into his next company: rapid experimentation, distinctive culture, and tackling messy real-world domains. The conversation closes with practical AI tool picks, a contrarian take on college, and lightning-round recommendations.
- •Startup principles: make many bets, test quickly, and ask for money early to validate
- •Culture as a competitive advantage: shared benchmarks for excellence and trust
- •AI tooling: Whisper Flow dictation, Claude Code as a terminal-based coding agent, Gemini model experimentation
- •Contrarian: college is underrated in tech; lightning round includes book/movie recommendations and ‘aim for Chartres’
