Transcription
Humanity is undergoing a massive transformation. More and more, the foundations of our future are being defined by data and, more and more, decisions based on the digital world are reshaping the physical one.
Today, gathering all the right data in one place is no easy task. Generating meaningful insights based on that data is even harder. And bringing those insights back to the physical world is harder still. But if the foundations of our future are defined by data... why are most companies still so bad at understanding it? And even worse at turning that understanding into meaningful results?
I'd like to share my future vision of a company called Palantir. A company that helps the biggest institutions make data a part of their foundation. Helps people make meaningful insights based on this new foundation. And, of course, leverage those insights to manage risks and make impactful changes in the real world.
To understand how Palantir can do this across every industry and discipline, we first need to talk about the challenges with big data, artificial intelligence, and the idea of a digital twin.
Hey there. My name is Alex, and I used to be a rocket scientist and radar engineer at MIT, where I focused on big data, artificial intelligence, and automated decision support for government programs—the exact kind of work that Palantir helps some of their biggest clients do at scale through their Gotham platform. Palantir has another platform, called Foundry, that's extending their reach to private institutions and even start-ups through their new Foundry for Builders program.
Because Palantir's technologies are industry-agnostic, I believe they can keep expanding downstream into new markets by eventually offering software products to more small and medium-sized businesses and maybe even individuals. So, in this episode, I'll cover the important things Palantir is doing today and how they could scale that to a trillion-dollar valuation in the far future.
If you enjoy this type of research and analysis, consider liking this video and subscribing to the channel with all notifications turned on. That way, you'll be the first to know when I come out with new research, regardless of how YouTube tunes its algorithm. Let's dive right into it.
Palantir builds advanced data analytics software for government and other large institutions. These same institutions tend to have the hardest time with big data at scale because they were built well before the world went digital. Palantir's software analyzes huge sets of data from many sources to help organizations make better decisions, often in near real-time.
So, where I think Palantir REALLY adds value is by helping these large, older companies quickly make what's called the "digital twin" of their organization. It's called that because it's a digital representation of their physical stuff. For example, the digital twin of a huge logistics company could look like a count of all of the assets it has at all of its locations, that updates when things are moved in the real world. Everything else Palantir provides depends on this digital twin, so Palantir really has to get this right.
To enable that, Palantir's software actually needs to solve several key big data challenges, EACH of which is big enough to build entire companies around. Let me walk you through them.
The first step is bringing all of a client's data to the right place, or places. That means working with the client to identify the right data sources, build the right integrations, and transport all of the data from those sources to the right storage locations. Initially, that could scare investors because it makes Palantir sound like a consulting company, but Palantir has done a great job of making this mostly plug-and-play to the point where new clients can be stood up in a matter of days. That's one reason their clients love them; they reduce the time to value from years to hours.
There are serious challenges with this step because privileged or classified data needs to be stored in accredited spaces, can only be transported through certain channels, and accessed by certain people—all of which can change over time.
By the way, that challenge isn't unique to the government; think about medical records, financial records, or even just protected programs that big companies don't want every employee to know about or have access to. So, if only some people in an organization can or should have access to this data, that company has to be able to manage access and permissions to all the different pieces of it.
In addition to that, data protocols and formats are always changing, so Palantir needs to be able to make secure fixes and updates to these connections over time. That's one of the main features of their Apollo platform, which is the continuous integration and delivery software that powers Foundry and Gotham. Apollo can deliver automated updates without disrupting operations.
Being able to roll out these software updates through even the most secure channels to the most locked-down computers is why Palantir can power mission-critical operations. That's a pretty big differentiator because, as it turns out, not many companies can actually do this reliably and at scale.
Okay, now that a client has the right data flowing to the right places through the right pipes with the right permissions, that data needs to go through ETL, which stands for extraction, transformation, and loading. Data needs to get checked for errors and inconsistencies, cleaned up and reformatted. It also needs to be made searchable, usable in models, and so on.
Derived data might need to be created, like calculating total revenue from sales quantities and prices or calculating energy from measurements of power and time. If it's privileged data, it might need to get anonymized or encoded or tagged a certain way, or maybe the system needs to generate a separate audit log alongside the data. The devil is in the details here, and these details vary company by company.
Then, that extracted and transformed data needs to be stored somewhere for the rest of the software to load it and act on it. That's called data warehousing, and every big company also has different data warehousing requirements based on their corporate culture, industry accreditation and compliance requirements, and other constraints imposed by their own clients.
More details, more devils. Remember, Palantir is doing this for some of the biggest enterprises on the planet, so they're pros at navigating through all of these extra hoops. Palantir deserves our attention as investors exactly because they automate these steps of data connection, aggregation, transformation, and storage in a secure, compliant, industry-agnostic way to help companies make data a part of their foundation.
Palantir helps people make insights based on this new foundation through all the other cool decision support tools their platforms offer: search, analytics, modeling and simulation, artificial intelligence, visualization, reporting, and so on.
Here's a quick example of how Palantir brings this all together to empower clean energy companies: There are plenty of companies that do data analytics, visualization, and reporting, similar to the dashboards that Palantir builds. Instead, let's stay focused on the capabilities I think really set Palantir apart.
We already talked about their ability to rapidly digitize their clients' business to build and interact with its digital twin. One way to do that is by running data through models and simulations. In the wind turbine example, the company might want to look at the effects of different weather conditions. So, they can do something as simple as run a simulation of a single turbine through different wind speeds and directions, or something as complex as hook their entire digital twin up to a global weather model and analyze how different weather conditions impact the power production of all their turbines together.
They can then steer the turbines in their digital twin to simulate the best angle for turbines and see the difference in power generation. The first notable thing about this is how easy Palantir makes it to do everything I just said due to their modular design and good graphic interface.
In my opinion, good user interfaces are kind of like the SATs. Remember those? Nobody cares if you have a good user interface; that's what's expected. Everyone cares if you have a bad one. It's just a box that Palantir needs to check, and they do check it.
The more impressive thing is that this connection between digital and physical is bi-directional, meaning if the simulation finds a better way to angle the turbine blades, a user can send that information directly to the turbine control system and move them in real life, or just output the set of instructions for a human to verify before doing it themselves.
This ability to hook into external systems, whether they're command and control systems for wind turbines or drones, or communications like emails and texts, allows these digital systems to affect the physical world. Think about how confident a client has to be in Palantir's software to allow it to connect directly to their wind turbine control system.
There are a few more capabilities I want to quickly point out before talking about how Palantir can expand into small businesses and the consumer market over the next decade. First, they have out-of-the-box use cases, called archetypes. The archetypes allow specific Palantir projects to get up and running fast by deploying a template that gets you most of the way there from the get-go.
Palantir is constantly expanding their set of out-of-the-box solutions with the types of use cases they're solving over and over again across their clients. This is huge for winning clients, because Palantir can walk in and demonstrate real value add in a matter of hours.
Second, Palantir supports third-party models. What this means is someone, whether it's an academic institution, research lab, or one of Palantir's customers, could make a better wind model that Palantir can then deploy to everyone that wants it.
When I worked at MIT, I got to work alongside engineers at MathWorks, the creators of MATLAB, on this same type of stuff. We would help them refine different signal processing models and functions, and then they'd update that package for all of their clients. When they worked with engineers at Stanford on their machine learning code and pushed an update, I benefitted from that as well.
Whether you're a utility company managing wind turbines or a government agency managing airborne drones, you always want access to the latest and greatest wind models, even if they're not made by Palantir. In this way, Palantir kind of acts as connective tissue across its clients, allowing an improvement from one company to propagate everywhere else that benefits from it, just like MATLAB did for me at MIT.
Speaking of Palantir acting as connective tissue, the third capability is Foundry's ability to create collaboration spaces BETWEEN companies. Here's the head of the Foundry program explaining how Palantir is thinking about this problem: Palantir's ability to create secure collaboration spaces across organizations is huge. Remember, we're not talking about emails and phones here. We're talking about organizations with very strict policies on information security, limited access data, legal and compliance standards, and so on.
The more enterprise clients Palantir gets on their platform, the more I think will want to join it for this reason alone. Palantir is often criticized for relying on government contracts, but I think this is a mistake. Thinking about Palantir's platforms as this connective tissue between organizations means shareholders should want them to penetrate as much of the government sector as they can.
Not just to lock in those long-term, stable contracts, but also to access models, tools, and processes that most other companies simply can't access. To me, that means Palantir benefits from some serious network effects, and we want them to fill that network with as many hard use cases as possible upfront.
With that said, let's talk about Palantir's future and how they could expand beyond their government and enterprise clients to small business and individuals. Palantir recently announced that it will roll out Foundry for builders, which is an initiative to get early-stage companies to use the Foundry platform as a subscription service. Because of the way Palantir can help a company quickly build its digital twin, I don't really see it having too much competition in this space.
Most other solutions either focus exclusively on one aspect of business, like customer relationship management or workforce management, or they only cover a small piece of Palantir's value chain, like only focusing on data visualization and reporting. Palantir being able to offer an end-to-end solution for many different types of businesses makes it a compelling option for any company that wants to do as much as they can with as few different ecosystems and tools as possible.
Spoiler alert: that's every company, which means this is a real moat for Palantir. Also, once a company builds on top of Palantir, they'll be very hesitant to move away from it because they'll probably have to find multiple expensive pieces of software to replace it. Palantir will focus on startups with high growth potential, starting with companies already connected to their existing clients, then move to the broader market.
I think this idea of thinking about Palantir as connective tissue between organizations really has some merit, and I'm really excited to see how their ecosystem of archetypes and models continues to evolve.
Now let's talk about the ultimate market. What could a company like Palantir ever offer an individual like you and me? Well, what if it treated meaningful collections of individuals as an organization? For example, what if Palantir came out with an internet browser that learns all about you, pools you with users like you, and recommends content and commerce based on what's going on in your pool of people?
It could be like Amazon, but for everything. Hey, people in your area are stocking up on toilet paper and soap. Here are the closest stores with some left, or a link to order some online. We think this may be linked to the following local news items. Here are other recommendations for things to read and do, based on these insights.
As you do different things, like book travel or buy a car or upgrade your phone, Palantir could dynamically adjust which pools of people you belong to and the types of recommendations it would make. What if Palantir integrated with a lot of other services and provided you with a universal login, kind of like how Facebook and Google do today?
Then, when you log in with your Palantir account, it adds that data to your data pool to make cross-service recommendations. Hey, people who liked this show on Amazon Prime really love this movie on Netflix. Hey, people who downloaded the fitness app you just got end up listening to these 4 playlists on Spotify more than any others. Hey, people who subscribe to Ticker Symbol: YOU tend to be really, really, ridiculously good-looking. I'd pay a buck or two a month for that.
My point is, we will never know what kind of large-scale, cross-platform pattern of life insights could be unlocked unless there was a company that could aggregate all of our data together and compare it to everyone else's, which Palantir is very capable of doing. Also, if there's one company I'd trust to keep all my data safe, it's Palantir.
Comment below with your thoughts on these ideas. Would you subscribe to a service like that for a couple bucks a month? Do you have your own ideas for services Palantir could offer to individuals in the next 5 to 10 years? I'm excited to hear your thoughts.
Palantir's next earnings call is only a few weeks away, so I'll do a follow-up episode on their finances, how their business units are growing, the crazy amount of SPACs they're invested in, and all of the other things we should go over when we focus on Palantir's stock.
Comment below or tweet me at ticker symbol you if you want me to cover anything else in that video, including Palantir's founders or more about their platforms. Likewise, I hope this episode helped you understand more about what Palantir does today and their path to expanding in new markets in the future.
If it did, let me know by investing in the like button and subscribing to the channel with all notifications turned on. Then you'll be notified as soon as my next Palantir video drops AND it's a great way to invest in the channel that invests in you.
Until next time, this is ticker symbol you. My name is Alex, reminding you that the best investment you can make is in you.