Palantir Technologies (PLTR)

Data is on an Exponential Rise

The amount of data produced per week, month, or year, is on an exponential rise. Yet the large majority of orgs still don't have the means of processing and holistically using their data to generate better business outcomes. At present, as orgs generate more data, the difference between their potential and reality is widening. Hence, the orgs that are implementing a modern data stack have a real chance of beating the competition.


Palantir's objective is for Foundry to become the OS for the modern enterprise, and be that connective tissue within the modern data stack. Foundry leverages SDDI (Software-Defined Data Integration) to connect to a wide range of data sources - legacy systems, business applications, SaaS apps, IoT devices, etc. It can also connect to cleaned and prepared data located in data lakes, data warehouses, and data lakehouses.

After ingesting the data and making any necessary transformations, Foundry creates an ontology so that users can visualise and understand all the relationships between data items. Palantir's ability to create customised ontologies for every customer is a key technological advantage they have over the competition.

Ultimately, the reason orgs have data siloes is because they lack a company-wide ontology that connects their digital assets in a shared vocabulary. Ontologies have made the Internet what it is today, encoding knowledge in web pages in a way that is searchable by electronic agents (e.g., Google). They are also critical in industries like healthcare, codifying diseases so IT systems across doctors, pharmacies, and insurance companies, can communicate with one another.

However, single companies don't have good internal ontologies because they are really hard to do. The exceptions are big tech names, though they have only created an ontology for themselves. Palantir has developed a capability to create a customised ontology for every customer - in a scaled way.  

Ontologies created by Foundry break down the data barriers to serve as the foundation to build better applications, perform better analytics, conduct machine learning, and generate richer insights to aid better decision making. Responses to outputs are fed back into the ontology to iteratively improve the understandings between data items and between inputs and outputs.

The other key advantage is that Foundry has been designed to accelerate the work of both technical and non-technical users, with a blend of code, low-code, and no-code platforms. Developers have the choice of a wide variety of programming languages, data scientists can choose from a plethora of ML models, and non-technical business users can build customised applications.


The other important aspect of Palantir's long-term strategy is Apollo. Initially, Apollo was created for internal use, as a DevOps automation tool to continuously deploy speedy software updates to Foundry and Gotham. In 2021, Palantir made Apollo available to customers wanting to deploy their own software.

In a digital world of increasing fragmentation and complexity, deploying the same software in multiple environments is extremely cumbersome. Apollo abstracts away the complexities, allowing customers to easily deploy their software to any environment (public cloud, private cloud, on-prem, edge, IoT, etc.).

Together, Foundry as the OS and Apollo as the software deployment for the modern enterprise, have the potential to transform how data is ingested and analysed, how applications are built, and how software is deployed. With these platforms, Palantir has more PaaS and SaaS services than AWS. The breadth and depth is unparalleled, yet the platforms are also highly interoperable with third-party OS, software, and environments.


If the execution is on point, then these attributes ought to make Palantir one of the most important supercloud, or sky compute, pioneers. Supercloud or sky, are the terms increasingly used to describe a vendor that can situate themselves atop multicloud environments as a next-gen abstraction layer, in order to reduce costs, simplify operations, and speed up the time-to-market for enterprises. Palantir has the potential to be the leader in this emerging class of software vendors.

We surmise that Palantir has a multi-year lead on their competition, mainly because there are no shortcuts for building software with such density. Palantir's software today is a culmination of many years whereby engineers literally sat side-by-side with end users to design the software they would be using. Over time, Palantir took these learnings to standardise features and functionalities, which eventually evolved into Gotham, Foundry, and then Apollo.

Execution Risks

There are substantial risks to the thesis, however. As investors ourselves, we're not alone in thinking they could accelerate the GTM success if they made Foundry more accessible to the wider market. They have recently become more developer-friendly, introducing demos and technical documentation. Though, to achieve a robust bottom-up GTM, they need to do more. And despite the technological superiority, if they don't improve the GTM traction then they will not maximise their market potential and lose out to the competition.

Attractive Unit Economics & TAM

Assuming Palantir does improve the GTM traction, over the long-term the business will likely become very profitable. The gross margin is ~80%, indicating attractive unit economics. S&M, R&D, and SBC as percentages of revenue are steadily declining, pushing EBIT toward profitability. FCF margin is positive and will trend higher as expenses mature and the business benefits from being a capital-light operation.

Lastly, Palantir has a huge TAM. Potentially thousands of enterprises or even SMBs could eventually be customers of Foundry. Furthermore, Apollo creates a whole other chunk of TAM that is difficult to measure - as every single org undergoing some form of digital transformation requires their software to be deployed to one or multiple environments. And given Palantir's first principles understanding in designing software, feasibly they could enter any new software market they pleased if they wished to expand the TAM some more.

For institutional investors, on request, we can do tailored research or memos into any specific aspect of Palantir. For all types of investors, here are individual reports either solely focused on Palantir or including Palantir to some degree.

For institutional inquiries, or to pay for individual (a' la carte) reports, please click Subscribe for more information. From there you can also sign up as Premium subscriber if you wish to.


Palantir: Parallels With Microsoft, Potential Network Effects Underappreciated (January 2021) - Free

Palantir Follow-Up: Getting To The Core Of The Competitive Edge (November 2021) - $50

Superclouds Part 1 (February 2022) - $50

Palantir Stock: An Emerging Supercloud With Attractive Valuation (March 2022) - Free

Notes & Framework For A PLTR Presentation (May 2022) - $50

Can Palantir Replicate The Success Of Intel, Microsoft, Apple, And Amazon? (May 2022) - Free

Palantir SBC - Let's Put This To Bed (May 2022) - Free

Palantir Vs. Bright Future Vs. Red Flags (July 2022) - Free

Cycles Of Innovation (June 2022) - Free (presentation on YouTube)

Tailored research requests - price negotiable