Recent articles in case you missed them:

Palantir is a software company that helps customers get the most out of their structured and unstructured data. Like many others, it pulls from years of AI/ML work to automate insight-gleaning. It utilizes complex neural networks to power anomaly detection, trend forecasting and natural language processing. It works openly with many database vendors for interoperable storage and low-latency querying, and offers its own tools there too.

Overall, PLTR frees clients to conjoin disparate data sources while utilizing its software to uncover ideas that manual analytics and legacy competition cannot derive. It gives customers a birds-eye view of their operations, with detailed suggestions to help optimize products and workflows. It also lets companies freely test digital twins (a process called Ontology) in zero-stakes environments to actually understand what works and what doesn’t. It’s a similar idea to split testing, with deeper, more applicable and more outcome-driven learnings.

Revenue is neatly split into two buckets – “government” and “commercial.” Government clients use its Gotham product platform, while commercial clients use its Foundry product platform. With Gotham, Palantir routinely builds custom use cases for individual government clients. Foundry was built to be more malleable, with more pre-built app integrations and developer kits available. That diminishes the need to conduct custom builds for every single private enterprise. Still, it does materially more custom building than a typical B2B software firm will.

It also seamlessly leverages the commercial platform to cater to industry-specific needs. By-industry models are intuitively named “micro-models.” These are smaller and boast sector-specific use cases with specialized regulatory compliance help. A financial services model from Palantir, for example, may assess credit risk or fraud detection.

Palantir Apollo provides continuous integration and continuous delivery (CI/CD) to automate software package building and deployment. It’s a foundational piece of the firm’s ability to collect, utilize and drive value from broad data ingestion. It’s also how Palantir can help operationalize these learnings to introduce valuable products.

In the realm of GenAI and Agentic AI, Palantir is not playing the game of building the biggest model or buying the most GPUs to have the largest infrastructure footprint. This is not a hyperscaler like AWS. It simply gives clients the tools and integrations needed to build apps for their own, more granular use cases. It also has some products for model customization, but the app layer is where this company really shines.

Its most exciting product here is called Artificial Intelligence Platform (AIP). This powers automation and AI app-building.  The company compares AIP to what public cloud vendors did for compute and workload modernization. AWS, Azure, Google and Oracle provided the environment, tools, storage, security and maintenance needed to grow needed compute capacity without managing it yourself. This made migrations and adoption the rational decision. AIP attempts to do the same thing in terms of pushing enterprises to adopt GenAI. This fully manages expedited product building for client deployment. It allows for open collaboration between software developers, data scientists, models and project managers to ensure effective work. It directly supports Foundry and Gotham by uplifting and augmenting potential use cases to “extract value from GenAI models.” And it does so in a quite compelling way that can craft tools on a by-customer basis.

As leadership will tell you, AIP isn’t just another dime-per-dozen chatbot. It’s an aggregator of data, tools and services needed to actually build valuable apps and to embrace GenAI. Considering the lack of finite and structured end products stemming from AIP, I think it helps to hear about some examples of what clients are doing with it:

The list goes on and on. AIP is where jumbled data, processes and ideas turn into the operationalized, actionable creation of GenAI products.

“We are still in the earliest stages, the beginning of the first act, of a revolution that will play out over years and decades… This is not an incremental advance or marginal acceleration of our business. This is a new phase… The business we have built has now developed its own internal momentum and strength… with the output that we are seeing far surpassing what we are investing. A software juggernaut has indeed emerged. .. We have the products and reach of an established incumbent and the speed, growth, and agility of an insurgent startup.”

Co-Founder/CEO Alex Karp

The mix of RPO, TCV, RDV and customer count outperformance uniformly depicts strong forward-looking demand trends enduring or accelerating further. This also marks the 6th straight quarter of Y/Y revenue growth acceleration for the company. Very impressive execution.

Palantir trades for about 125× 2025 EBIT using its guidance and not lower analyst estimates that will surely rise following this report. Current 2-year EBIT growth estimates call for a 26% CAGR, but I think we can safely call that 30% following this showing. This gives it an EBIT growth multiple of over 4.0x.

AIP – Thriving & DeepSeek:

As leadership puts it, while everyone has been focused on model improvement and differentiation there, Palantir has been laser-focused elsewhere. It does not care who builds the best model… it will simply use whatever is best. And? As the tools it uses become cheaper and cheaper, its operations directly become more efficient. DeepSeek is uniformly good news for this company. Palantir called the vendor’s new model impressive, but also said they “absolutely did steal a lot of the work through model distillation and might have stolen even more.” It also thinks they had access to far more compute to build this model than many think, citing massive, over-indexing Nvidia orders from Singapore as evidence.

For Palantir, the firm has fixated on using this increasingly similar base of large language models (LLMs) to differentiate via data and excellent software releases such as AIP. “Excellent” is a subjective term, but when digging into the performance here, it’s very difficult to argue with that classification. While it doesn’t split out performance here in isolation, adoption can be seen within its thriving U.S. commercial segment. Remaining deal value rose 99% Y/Y and it closed a new record for TCV booked during the quarter. Incredibly, TCV crossed $800 million this quarter, which nearly doubled the previous record. 

Deals rose 45% Y/Y, which points to sustainably strong demand ahead, as customers routinely land and aggressively expand with this business. That can be seen in continued net revenue retention (NRR) improvement, which rose to 120% Y/Y.  Many, many software companies are struggling with this specific metric, as their businesses shift to new logo generation vs. existing expansion. Palantir is having that same great new logo success, as the number of deals closed for this quarter represented 50% of its total U.S. commercial customer count. It’s just overcoming this welcomed NRR headwind via fantastic existing client momentum too.

Leadership sees “significant expansion opportunities” with these customers. For its largest 20 customers, revenue is still growing 18% Y/Y to an average of $65 million per logo per year. AIP is helping a ton.

AIP – Where Momentum is Coming From & Case Studies:

It’s all working for AIP. Why? I’m glad you asked. Leadership is often outspoken about its unmatched ability to actually build value-creating software in the world of GenAI and Agentic AI. As they’ll tell you… competition has slide decks and chatbots, with Palantir having a boatload of utility-driving tools (some via software development kits and some via custom builds) to conduct more complex tasks.

To the team, this stems from the culmination of decades of work in building ontology systems to unlock rapid, data-driven testing. Palantir CTO Shyam Sankar describes ontology as providing an “intermediate representation of the enterprise” so that AI can interact with something much closer to actual operations. This sharpens the learning curve. Ontology pushes Palantir towards its vision of helping clients build “autonomous companies” that are supervised by agents and people, rather than firms where tasks are largely manually completed.

This foundation means expedited insight gleaning and product improvement, without conceding operational disruption and while greatly diminishing the risk of failed product launches. Simply put, with Palantir, companies can understand exactly what piece of automation and what workflow works. They can do this within a secure, zero-stakes environment before deploying all of this to runtime. Palantir also gives customers significant guidance and tools to help them know what to build for specific problems.

It’s this dynamic that leads to, it thinks, superior products. While that’s nice to hear, it’s more real when coinciding with elite results and case studies, like for Palantir. An American telecom company signed a $40 million expansion deal to help them sunset older network technology and “achieve significant cost savings.” Panasonic Energy North America is using AIP to automate pieces of work for 350 technicians, thus “reducing machine downtime and improving throughput. A large pharmacy company used AIP to automate script load balancing and patient outreach workflows. After testing out these products, it signed a large $67 million contract thanks to the unmatched efficacy. Palantir helped an auto supply company automate a 100-hour process and also helped a large bank automate its core operations to cut some task time from days to minutes. Tangible value… being delivered today. That’s why AIP is winning.

Looking ahead, Ryan Taylor (Chief Revenue Officer) told investors the conversations with customers center around how they can do more with AIP. I’m sure Palantir is happy to help.

“It's a substantial leap to deploy LLMs into production with real impact. Most organizations are currently stuck on the wrong side of the widening chasm, working on their 2, 5 & 10-year plans, which become obsolete days later. Not Palantir… We've taken our position through our decades-long investment in developing the ontology, which allows organizations to seamlessly weave LLMs into their enterprise, unlocking their highest potential leverage. In this AI revolution, anyone looking for a solution that actually works is going to choose Palantir.”

Chief Revenue Officer Ryan Taylor

“We are convinced the normative value for AI is enterprise autonomy, the self-driving company. Users go from performing the workflow to supervising an army of agents, teaching them how to handle edge cases and reducing dwell time. This is where we are maniacally focused with our customers.”

CTO Shyam Sankar

AIP Go-To-Market:

While the external sales team is now being scaled here, a large portion of the selling activity so far was done via bootcamps. This is where it hosts events to provide hands-on support and “get clients from 0 to use case in 5 days.” All it needed to do from the start was host interested customers and demo AIP to enjoy this aggressive revenue ramp. And while that’s extremely convenient, there is much more business to be secured by going out and taking it, rather than relying on more inbound requests. This should be yet another growth lever for this thriving product. 

Warp Speed Re-Industrializing the USA:

Operation Warp Speed is a modern industrial operating system (OS)” that equips companies and governments with cutting-edge enterprise resource planning (ERP), product lifecycle management (PLM) and a manufacturing execution system (MES). It’s a fully managed way to rapidly allow manufacturers to fix how they build things. This is how Palantir plans to help “reindustrialize” the United States and ensure we build everything we need here. Early on, Panasonic Energy, L3 Harris, Shield AI and Anduril are using this OS to automate once-highly manual tasks. They’re leaning on Warpspeed for use cases such as automated inspections and scheduling too. Per Andural CIO Tom Bosco, the product has already delivered a 200X boost to its ability to “anticipate and respond to supply shortages.”

“The Warp Speed response has been exceptional and the pipeline is swelling.”

CTO Shyam Sankar

Government Wins From Around the Globe:

In the USA, the Army added 4 more years to its Army Vantage contract for Palantir to run its automated data analytics. Elsewhere in the USA, it signed its very first Mission Manager deployment with the U.S. Special Operations Command. As a reminder, Mission Manager automates the safe onboarding of new software and provides an overarching view of hygiene and permissions.

Palantir has invested heavily in things like Combined Joint All-Domain Command and Control (cJADC2), which is a military initiative meant to tear down operational silos and create better interoperability. Project Maven (the AI arm of cJADC2) continues to enjoy rapid adoption, and Palantir is a prime beneficiary of that trend. Interestingly, one thing Palantir provides is drone service for border security. Regardless of our thoughts on politics, the focus and budget allocated to that use case should rise over the next 4 years. Per Palantir leadership, they’re “just getting started on Maven.”

Leadership is also quite excited about the new Department of Government Efficiency (DOGE):

“Palantir's real competition is a lack of accountability in government. It's these forever software projects that cost an insane amount and don't actually deliver results. And I think DOGE is going to bring meritocracy and transparency to government, and that's exactly what our commercial business is. The commercial market is meritocratic and transparent, and you see the results that we have in that sort of environment.”

CTO Shyam Sankar

Across the pond, 26% Q/Q international government growth was thanks to its National Health Service (NHS) contract in the UK. 87 hospitals and 28 care provision providers are using its Federated Data Platform (FDP) for analytics and decision-making. In the rest of Europe, leadership continued to be highly critical of regulators and lamented on the continent “looking to the past as a way of getting to the future.”

FedStart:

As a reminder, FedStart is Palantir’s accreditation program for FedRAMP certifications needed to sell software to the government. It shrinks the time and cost it takes to secure this status for others and makes Palantir more of an ally vs. an enemy for more software companies. This quarter, FedStart secured approval for its FedRAMP high product, which is the highest level of authorization for the most sensitive data and needs.

International Commercial Business Notes:

Another masterful quarter from a company monetizing GenAI software better than anyone else on the planet. AIP has been the best software launch of the last few years – period. Continued accelerated growth paired with rapid operating leverage is the byproduct, and they still haven’t even built out an external sales team yet. Guidance is fantastic and forward-looking demand metrics point to those expectations being prudent, if anything. It is wildly difficult to overstate how ridiculously impressive these results are.

Now for the elephant in the room. The company is expensive. Really expensive. This elite report will surely lead to upward profit revisions that make the multiple look a bit more reasonable. But it will still remain very expensive. Assuming a vastly outperforming 40% 2-year EBIT CAGR from today still gives it a growth multiple of over 3x and still makes this the most expensive name that I cover. Does that mean it cannot keep working? No. It simply means that the price tag demands future quarters look as amazing as this one did. Future reports will need to drive large incremental upward profit estimate revisions. That’s what bulls are banking on… that’s what Palantir delivered today.

I don’t see the risk/reward as compelling here for the stock. I have no interest in owning it today (hopefully eventually). But I am deeply, deeply impressed by the company and think bulls should be taking a well-deserved bow. You have earned it and then some.