PALANTIR: War, Surveillance, Artificial Intelligence

Part 4: Digital Transformation of Enterprise and the Public Sector

Chapter 7. Opportunity in Crisis: Monetizing Chaos

Attorney Kyungjin Kim

A. Savior of the Pandemic

(1) Building the NHS Vaccine Distribution and Bed Management System in the UK

In March 2020, during the same week the Nightingale Hospital was being hastily assembled inside London's ExCeL Centre, data managers at NHS England were facing a nightmarish reality. Patient data pouring in from 215 hospital trusts across the country arrived in different formats, from different systems, at different speeds. Which hospital had ventilators to spare, which intensive care unit had an empty bed , nobody could see the full picture.

That's when a phone call came in. On the other end was Palantir's sales team, freshly arrived from the United States. They offered their software for just one pound. There was one condition: give them access to NHS data. Palantir's Foundry platform soon became the nervous system of the NHS vaccine rollout. When the Pfizer vaccine required storage at minus 70 degrees Celsius, the platform connected cold storage locations across the country, transport truck routes, and appointment schedules at each vaccination centre in real time. At the North Tees and Hartlepool Trust, long-stay patients hospitalized for 21 days or more dropped by 36%. In Dorset, 2,500 additional surgeries became possible per year. In North Cumbria, surgical throughput improved by 10%.

The numbers were impressive. But behind them lurked an uncomfortable question: the medical records of 65 million Britons were flowing through an American company's servers.

(2) Establishing Itself as Public Health Infrastructure and the 330-Million-Pound Contract

On November 21, 2023, NHS England signed a seven-year contract with Palantir worth 330 million pounds (approximately 550 billion Korean won). Named the 'Federated Data Platform' (FDP), it was the largest healthcare data contract in British history.

The contract ran to 586 pages. Then something strange happened. When civil society activists opened the contract published on the government register that December, they couldn't hide their bewilderment. 417 pages had been completely blacked out. Seventy-one percent of the entire document was redacted. The 'data protection' sections were almost entirely erased. Financial reporting content had vanished. The roles and responsibilities of key personnel were unreadable.

Ian Brown, a lawyer at the Good Law Project, was furious. "Redacting 70% of the contract is unlawful in itself. It directly violates government transparency guidelines." He immediately launched legal action, sending a pre-action protocol letter to NHS in February 2024 alongside openDemocracy and Foxglove.

An even more shocking revelation was waiting. Shortly after the contract was signed, emails leaked showing that Palantir had hired PR agency Topham Guerin to pay social media influencers to attack the Good Law Project. The influencers were instructed not to mention Palantir's name.

NHS England launched an investigation. The contract explicitly stated that Palantir needed prior written consent before using the NHS name or brand in marketing or promoting the contract. Barely weeks after signing, allegations of contract breach had already surfaced.

In the spring of 2024, FDP entered full operation after a six-month transition period from the existing national data platform. The goal was to connect 240 healthcare organizations by 2027. In a business case analysis published in October 2025, NHS projected a fivefold return on investment. But reality proved difficult. By the end of 2024, fewer than a quarter of 215 hospital trusts were actually using FDP. Even by May 2025, only 72 trusts , a third of the total , had adopted it. Greater Manchester health authorities wrote in an internal document: "None of the products designed or produced by Palantir as part of the FDP programme currently exceeds NHS Greater Manchester's own capabilities."

Leeds Teaching Hospitals NHS Trust was blunter. In a letter released through a freedom of information request, they wrote to NHS England: "Adopting some of Palantir's platform tools would mean losing functionality, not gaining it."

Eventually, the Department of Health paid KPMG 8 million pounds to 'accelerate' adoption of Palantir's software. A consulting firm having to sell another company's software , that was the state of British public health infrastructure in 2025.

The controversy was amplified by Peter Thiel's remarks. Palantir's co-founder and board chair gave a speech at the Oxford Union in 2024 where he said: "The NHS should be privatized. The British affection for the NHS is a form of Stockholm syndrome." The company distanced itself, calling it a statement made in a 'purely personal capacity.' But the fact that this company's founder was calling for the dismantling of the very public healthcare system whose data he managed was not easily forgotten.

A senior official at Amnesty International UK commented: "Palantir is a deeply concerning choice as an NHS service provider, given the company's reputation surrounding human rights controversies." He was referring to its work with U.S. Immigration and Customs Enforcement (ICE). A 2020 Amnesty International report had warned that Palantir posed a "high risk of contributing to serious human rights violations."

In January 2025, Palantir announced a partnership with Multiverse. The plan was to train NHS staff on using the data platform through an FDP apprenticeship programme. This was no ordinary training initiative. Once staff become accustomed to Palantir's system, the cost of switching to anything else grows exponentially. That is the lock-in effect.

The pandemic crisis opened the door to British public health for Palantir. It took three years for a one-pound contract to grow into a 550-billion-won deal. The alchemy of turning chaos into profit , Palantir had perfected it.

B. Supply Chain Collapse and Recovery

(1) How Global Companies Like Airbus (Skywise) and BP Used Foundry

Paris Air Show, 2017. Airbus's head of digital transformation took the stage to unveil an ambitious vision: connecting all data across the aviation industry into one system. Standing beside him was a Palantir engineer. It was the birth of a platform called 'Skywise.'

Skywise had a distinctive architecture. Palantir's Foundry provided the backend infrastructure, while Airbus layered aviation domain expertise on top.

A single aircraft generates roughly 500 gigabytes of data per day for an A330neo, and up to one terabyte for the latest A350. Engine sensors, flight records, maintenance histories, parts inventories, flight schedules , all of it flowed into a single platform.

Airbus opened this platform to airlines. For free. Skywise Core was included as a standard option with aircraft purchases. Buy a 100-million-dollar airplane, and it comes bundled at no extra charge. Airlines could compare their fleet's operational data against industry averages. But the real revenue came from premium services: Predictive Maintenance, Health Monitoring, Fleet Reliability analysis. Those required a subscription. Once airlines grew comfortable with the free version, they naturally graduated to paid tiers.

By 2023, more than 50% of the global Airbus fleet , over 11,900 aircraft , was connected to Skywise. More than 140 airlines had joined. EasyJet, AirAsia, Cathay Pacific, Philippine Airlines, Malaysia Airlines, and Delta Air Lines were all inside this ecosystem.

Behind the numbers were real stories.

Jetstar, the Australian low-cost carrier, used Skywise to detect system performance degradation in its A320 fleet before problems materialized. The results were striking: engineering-related cancellations dropped by 93%. Predictive maintenance prevented dozens of flight cancellations.

A dramatic case occurred in August 2022. A330neo operators reported high-pressure valve leaks in the engine bleed system. Airbus engineers analyzed operational and sensor data accumulated in Skywise. They discovered that under specific takeoff settings, the bleed monitoring computer's software failed to properly control the high-pressure valve. Excessive stress caused the valve clamping pins to fracture, leading to seal damage and leaks. In a worst-case scenario, high-temperature, high-pressure bleed air could seep into the wing and cause catastrophic structural damage.

Airbus immediately reported to the European Union Aviation Safety Agency (EASA) and recommended issuing an emergency airworthiness directive. They chose to accept the cost of grounding aircraft rather than risk a catastrophic accident. It was a case of big data preventing disaster.

Airbus's goal was clear: achieve 10 billion dollars in services revenue over the next decade. Skywise was the core engine driving that target. Independent analysts estimated the platform was generating over 850 million dollars in annual revenue opportunities.

The relationship between Palantir and Airbus resembled that between Veeva and Salesforce. Just as Veeva built life-sciences solutions on top of the Salesforce platform and paid 40 to 50 million dollars annually in royalties, Airbus built aviation-specific solutions on Foundry. Of the 26,900 Skywise users, the majority were Airbus employees. Fifteen percent of its 130,000-person workforce used the platform.

Meanwhile, in the middle of the cold North Sea, another digital twin was at work , on BP's offshore oil platforms.

In 2014, BP signed its first contract with Palantir. The wounds from the 2010 Deepwater Horizon explosion had not yet healed. Crude oil gushed for 87 days in the Gulf of Mexico , 7.8 million barrels. It was the worst marine oil spill in history. BP realized that digital transformation was no longer optional but a matter of survival.

Palantir built digital twins across BP's oil and gas production operations. North Sea offshore platforms, Gulf of Mexico deepwater drilling rigs, the Khazzan gas field in Oman. Data flowed in real time from more than two million sensors , pressure, temperature, flow rate, vibration. These numbers were transformed into digital replicas of physical assets.

In 2021, BP extended the contract by five years. Then in September 2024, the two companies agreed on a new level of collaboration: deploying Palantir's AIP (Artificial Intelligence Platform). AIP applied large language models (LLMs) to industrial operations. When a BP engineer asked in natural language, "Analyze the cause of pressure anomalies on Platform 3," AIP synthesized relevant sensor data, maintenance records, and similar past incidents to provide an answer. The critical feature was hallucination control. To prevent LLMs from confidently producing false information, transparency tools were applied to every AI recommendation. Security features constrained the LLM's range of action, and every decision and action was logged as a fully auditable digital record.

Matthew Babin, Palantir's head of energy and natural resources, said: "The goal going forward is the same , driving greater efficiency across BP operations and expanding data integration. Now AIP offers an opportunity to accelerate human decision-making on top of the solid digital twin and deep operational workflows already in place."

BP's subsidiary Castrol built a supply chain digital twin. Azule Energy, a joint venture between BP and Italy's Eni operating in Angola, also signed a multi-year contract with Palantir to manage daily production of 200,000 to 250,000 barrels. In August 2024, a collaboration with NASA was announced: BP's digital modeling technology, developed for drilling in hard-to-reach locations, would be applied to lunar and Mars exploration.

Airbus and BP. What did these two companies share? Both operate complex physical assets. Both depend on safety as a matter of life and death. Both were drowning in oceans of data. Palantir launched a ship on those waters. The price was steep. But without that ship, navigation itself seemed impossible.

(2) 'Warp Speed': A Manufacturing Operating System and American Reindustrialization

On December 11, 2024, Palantir announced a new program called 'Warp Speed.' It was an operating system for manufacturing. The first 'Cohort' included four companies: Anduril Industries, L3Harris, Panasonic Energy of North America, and Shield AI.

Palantir CTO Shyam Sankar declared in the announcement: "When World War II began, we didn't have a 'defense industrial base.' We had an 'American industrial base.' Our future must look the same. America must reindustrialize and mobilize at warp speed."

There was historical context behind that statement. When the United States entered World War II in 1941, automobile factories built tanks, and refrigerator factories made machine guns. Civilian industry converted to military production. Sankar was arguing that the same kind of national mobilization was needed again , only this time, software would make it possible.

The design philosophy was simple: "Software must adapt to the business. Not the other way around." Traditional enterprise software , SAP or Oracle ERP systems, for example , demanded that companies reshape their business processes to fit the software. Years of implementation, hundreds of millions in costs, and results that still didn't match what the floor actually needed. Warp Speed promised the opposite.

Testimonials from the first cohort members were striking.

Anduril Industries is a defense startup building autonomous weapons systems. They were the first company to adopt Warp Speed as part of their 'Arsenal OS' product line. Chief Information Officer Tom Bosco said: "Delivery speed is everything for us. Warp Speed lets us rapidly build manufacturing capacity for our customers. Using this software, we've seen up to a 200x efficiency improvement in our ability to anticipate and respond to supply shortages."

Two hundred times. It sounds like an exaggeration. But in context it makes sense. Dynamic material requirements planning (MRP), efficient phased introduction of new configurations, reduction in material and labor variances , tasks that once took days were now done in minutes.

Shield AI builds unmanned aerial systems. Their flagship product, the V-BAT drone, was facing record demand. CEO Ryan Tseng explained: "Our mission demands rigorous, integrated execution from design to delivery. Production is only as fast as the slowest part or process. Warp Speed will help our teams identify bottlenecks and stay in sync."

L3Harris announced a strategic partnership with Palantir in October 2024. This legacy defense company was using Warp Speed to close the loop between design, configuration, and production. David Jack, Vice President of Global Operations and Program Excellence, said: "Warp Speed is a key element of our strategy to improve operational efficiency. It enables improvements like process automation and smarter inventory management." Panasonic Energy of North America was transforming American battery manufacturing through a three-year contract. Warp Speed supported the ramp-up at the Nevada Gigafactory and the Kansas De Soto plant. The goals were strengthening U.S. energy self-sufficiency and building resilience against global supply chain disruptions.

Warp Speed offered four core capabilities: Dynamic Production Scheduling, Engineering Change Management, Automated Visual Inspection, and a continuous network planning function called MRP Speed.

Among these, Engineering Change Management applied the CI/CD (Continuous Integration/Continuous Deployment) concept from software development to manufacturing. Design changes instantly reflected on the production line , something hard to imagine in traditional manufacturing.

One line from Sankar's speech stood out: "What we're supporting are companies that manufacture core products underpinning freedom and prosperity." Anduril's drones, Shield AI's unmanned aircraft, L3Harris's communications equipment, Panasonic's batteries , all were essential to American security and economic self-reliance.

Warp Speed was not just manufacturing software. It was an ambitious project aiming to become the software foundation for American reindustrialization , countering China's manufacturing rise, overcoming supply chain vulnerabilities, and accelerating the growth of defense innovation companies. Palantir intended to be the operating system for all of it.

C. The Bootcamp Strategy

(1) A Unique Sales Approach: Automating AI Workflows in Five Days

In mid-2023, Palantir began an unusual sales strategy: inviting prospective customers to a five-day intensive workshop. They called it the 'AIP Bootcamp.' Participants brought their own real data. Palantir engineers built AI workflows directly on top of it. Five days later, attendees witnessed problems they hadn't solved in years get resolved in a matter of days. CEO Alex Karp explained the strategy's purpose on an earnings call: "The AIP Bootcamp is a challenge we throw down, daring customers to compare against what they've built internally. No matter how many resources they've poured in or how many years it took , run your data on Palantir's platform for just ten hours, and you'll see the difference."

The results showed up in the numbers. At the end of 2022, total bootcamps numbered just 92. By the end of 2023, the count had surged past 500, with over 465 organizations participating. By June 2024, the cumulative total exceeded 1,300.

One case illustrates the power of this approach. Palantir closed a 3-million-dollar contract through outbound sales. Then they ran a bootcamp. Before the same quarter ended, that contract expanded into an enterprise-wide deal , several times the original size.

CRO (Chief Revenue Officer) Ryan Taylor reported: "By the end of November, we plan to run bootcamps for more than 140 organizations. Nearly half of those will happen this month alone. That's more than all U.S. commercial pilots from last year combined."

The bootcamp's effects appeared across multiple dimensions.

First, it shortened the sales cycle. Traditional enterprise software sales typically take six to eighteen months: requirements analysis, proposal writing, technical validation, pilot projects, negotiations. The bootcamp compressed all of that into five days.

Second, it reduced customer acquisition cost (CAC). Instead of sales teams scheduling dozens of meetings to persuade potential buyers, the five-day intensive experience did that work. Seeing is believing. Executives who watched their own data being transformed were easily convinced.

Third, it created cross-sell opportunities. During bootcamps, participants were exposed to Palantir's various products: Foundry, AIP, Apollo. Many came to solve one problem and ended up purchasing multiple products simultaneously.

Fourth, it built internal capabilities. Palantir's blog explains: "Participants build their own intuition and skills so that in the weeks that follow, they can independently identify and execute AIP use cases." Palantir experts were being created inside customer organizations. These people became internal evangelists driving platform adoption.

Fifth, it strengthened the lock-in effect. Once workflows were built through a bootcamp, the cost of switching to another platform spiked. Data structures, business logic, and user training were all tailored to Palantir. In August 2024, a collaboration with Microsoft gave this strategy new momentum. Palantir's AIP could now be deployed in the Azure Government Secret environment. The two companies planned joint bootcamps for intelligence and military customers , Microsoft's government cloud infrastructure combined with Palantir's AI platform.

(2) The Virtuous Cycle of Customer Acquisition and Lock-In

The bootcamp strategy was directly reflected in financial performance.

In Q3 2025, Palantir's quarterly revenue reached approximately 1.18 billion dollars , a 39% increase year over year. U.S. revenue alone was 628 million dollars, up 55%. The U.S. commercial segment grew 65% year over year, the fastest-growing division.

Behind this growth were bootcamps. Analysts noted that bootcamps had replaced the traditional multi-month sales cycle. Customer acquisition costs plummeted, and profitability surged as a result. In the second half of 2025, Palantir's operating margin hit 51% , an unprecedented figure in the enterprise software industry.

There's a metric called the 'Rule of 40.' It sums revenue growth rate and profit margin , a measure of the balance between growth and profitability. Companies scoring above 40% are generally considered excellent. Palantir's Rule of 40 score was 114%. Management boasted it was the second highest among major tech companies, trailing only Nvidia.

The bootcamp strategy soon became a benchmark for competitors. Snowflake launched its own 'LLM Bootcamps.' But replicating Palantir's approach , building workflows on-site with a customer's actual data , was not easy to copy. C3.ai was assessed as losing market share to Palantir's bootcamp model due to its longer deployment timelines.

CTO Shyam Sankar explained the essence of the bootcamp: "The AIP Bootcamp drives home the point that LLMs alone are not enough. You need tools that provide algorithmic reasoning." He was talking about the ontology , Palantir's core technology that transforms data from mere streams of numbers into 'digital twins' of the real world. The bootcamp was the device that let people experience its power firsthand.

In the second half of 2025, Snowflake and Palantir forged a strategic partnership. Two companies that had once been fierce rivals joined forces. Palantir's AIP could now run natively on Snowflake's data cloud. Snowflake handled data storage; Palantir handled the 'operational logic.'

This partnership opened a new land-and-expand channel for Palantir , the ability to spread through Snowflake's massive customer base.

CEO Karp said commercial success was no surprise. "Given our military experience, it's only natural." He had a point. For 20 years, Palantir had supported decision-making in the fog of war. That experience had simply been transferred to the corporate battlefield.

"AIP lets companies manage LLMs and essentially run penetration tests on their own enterprise." This remark from Karp revealed AIP's true nature. It wasn't about deploying an AI chatbot. It was about deeply integrating AI across an enterprise's data and processes , and ensuring that integration remained safe.

On an earnings call, Karp declared: "My view of what we need to do is to build a product so good that competition stops. Whether in commerce or on the battlefield. That's what we're doing. That's what we're seeing with AIP."

In 2025, Palantir emerged as the biggest beneficiary of the 'AI supercycle.' Companies began demanding real ROI from their AI investments, and Palantir could prove that value in five days.

A simple idea , the bootcamp , became the growth engine powering a company worth hundreds of billions in market cap. Palantir's ability to turn crisis into opportunity didn't come from technology alone. It came from precisely identifying a customer's urgent need and presenting a solution in the fastest way possible. And once that solution took root, pulling it out was nearly impossible.