🥛 Software is bleeding, who wins? 💻

GM. This is Milk Road AI, the only inbox you need when markets can’t decide whether they’re rotating or reinventing themselves.

Here’s what we’ve got for you today:

✍️ The only software company getting stronger?
🎙️ The Milk Road AI Show: Why Your First Humanoid Robot Won’t Be at Home... It’ll Be at Work.
🍪 Apple’s next iPhone update is bigger than it looks?

Okara’s Reddit agent finds high-intent threads 24/7 and writes on-brand comments that convert. Grow your business with Okara's Reddit Agent.

Prices as of 10:00 a.m. ET.

THE SOFTWARE BLOODBATH HAS A WINNER?

In 1453, Constantinople was the most fortified city on Earth.

It had the Theodosian Walls, a triple-layered stone-and-brick wall that had survived every siege for over a thousand years.

Arabs, Vikings, and others all tried and failed.

Then a Hungarian engineer named Orban showed up at the Ottoman camp.

He didn’t bring a bigger battering ram, longer ladders, or more soldiers.

He brought a 27-foot-long bronze cannon called the "Basilica" that could hurl a 1,200-pound stone ball over a mile.

For the first time in history, walls didn't matter.

The most legendary defenses in human civilization were reduced to expensive rubble, and Constantinople fell in 53 days.

Source: Makeagif

The enterprise software industry just had its own Constantinople moment.

And one company is holding the cannon.

But before I name the cannon, we need to understand what’s collapsing.

Right now, that collapse is playing out across the software sector.

J.P. Morgan called it the "largest non-recessionary 12-month drawdown in over 30 years”.

And roughly $2T has been erased from software stocks, cutting the sector’s weight in the S&P 500 from 12.0% to 8.4%.

This is a structural repricing of an entire industry.

For 20 years, these companies built empires on the same equation:

More employees = more seats = more revenue.

Salesforce alone built a $30B annual revenue business on the idea that every salesperson needs a "seat" to manage their pipeline.

ServiceNow did it for IT, Adobe did it for creative tools.

But here's the problem.

If an AI agent can run your back office, write your campaigns, manage your sales funnel, analyze your data, and spin up better software overnight:

Why would you pay $200 a month for every employee to use legacy software?

The answer is: you wouldn't.

And investors just figured that out all at once.

But to be fair, some companies are facing real disruption, and others are simply guilty by association.

And while every other software company is scrambling to tape bricks back together, one company is sitting there calmly, polishing its cannon.

It’s none other than Palantir.

Inside Palantir’s business model

Palantir was founded in 2003 by Peter Thiel (PayPal co-founder), Alex Karp (a philosopher with a doctorate in neoclassical social theory, yes, really), and a small group of PayPal veterans and other engineers.

The founding thesis was born on September 11th.

The U.S. intelligence community had mountains of data, satellite imagery, intercepted communications, and financial records, but it was all locked in separate databases that couldn't talk to each other.

The CIA had one system, the FBI had another, the NSA had a third, and nobody could connect the dots.

Thiel and Karp realized that the fraud detection tech built at PayPal, software that spots suspicious patterns across massive datasets, could be adapted for national security.

So they pitched it to Silicon Valley, and they all laughed.

The only investor willing to take the bet was In-Q-Tel, the CIA’s venture capital arm (a name that sounds like it should be on a watchlist, not a cap table), which put in $2M.

Thiel bankrolled the initial $30M himself.

It took two full years of continuous iteration for the product to be ready.

By 2009, Palantir had become the connective tissue linking the CIA, FBI, NSA, DHS, and dozens of other agencies.

Fast forward to today: $315B market cap, 4,100+ employees, and four core platforms serving government and commercial customers across 40+ industries.

Let’s break down the machinery behind the empire.

First up is Gotham, Palantir’s core defense and intelligence platform, released in 2008.

It takes data from every source you can imagine, satellite imagery, drone feeds, intercepted comms, financial transactions, sensor data, classified databases, and integrates it into a single, searchable platform.

Source: Harvard

The CIA, NSA, FBI, Department of Defense, UK Ministry of Defence, the Ukrainian military, NATO allies, and many other organizations use it.

U.S. forces used Gotham to predict the locations of roadside bombs in Afghanistan, helping save lives.

Palantir also took over the Pentagon's Project Maven AI contract (after Google declined) and secured a $1.3B deal running through 2029.

And here's the kicker: Gotham is authorized as one of only five SaaS offerings for Mission Critical National Security Systems by the Department of Defense.

That authorization alone is a moat that would take a competitor a decade to replicate.

Next is Foundry, Palantir’s commercial platform that applies the same battle-tested architecture from defense environments to enterprise problems.

It creates a "digital twin" of an entire organization, a live, dynamic model connecting data to real-world assets like factories, supply chains, equipment, and people.

Airbus used it to quadruple production of the A350 by merging 25 data silos.
Scuderia Ferrari uses it for split-second pit decisions during Formula One races.

Foundry is the product driving Palantir's explosive commercial growth, 137% year-over-year in Q4 2025 in the U.S. alone.

Next is Apollo, Palantir’s deployment and delivery system, which most people overlook but is a major reason the software becomes so hard to replace once installed.

It's the deployment and update system that works across cloud, on-premises, classified networks, battlefields, remote oil rigs, and even naval vessels with spotty internet.

Once a client deploys through Apollo, the switching costs become enormous.

Ripping it out would be like performing open-heart surgery on your data architecture.

Next is AIP, Palantir’s newest platform, launched in April 2023, which has driven much of the recent excitement around the company.

It integrates large language models (the tech behind ChatGPT and Claude) into secure, private enterprise environments.

But here's the critical difference from what everyone else is doing:

AIP doesn't just bolt a chatbot onto your data. It connects AI to actual operations.

Organizations can deploy LLMs on their own private networks.

Build custom AI agents through a visual interface, no PhD required.

Implement guardrails that prevent hallucinations.

Create agents that don't just read data but actually take actions (updating databases, triggering workflows, writing back to operational systems in real time).

And the go-to-market is brilliant:

Instead of the typical 6-12 month enterprise sales cycle, Palantir runs "AIP Bootcamps", intensive 1-5 day sessions where a customer's team builds a working AI use case on their actual data.

You leave with a working prototype, not a PowerPoint deck.

By 2025, the bootcamp model had become the primary engine of their commercial growth.

But wait, there’s more, and this is the secret sauce no one else can copy.

HOW TO GET YOUR FIRST 100 PAID USERS ON REDDIT

Most crypto projects grow their customer base on X.

But one of the most underutilized growth channels is Reddit.

That’s exactly the type of problem the Okara Reddit Agent is built to solve.

Here’s what you have to do:

Go to the Okara Reddit Agent dashboard.
Drop in your product URL and website.
Add the keywords and topics your ideal customers care about.

The Okara Reddit Agent then works 24/7 to:

Monitor thousands of communities
Find threads where your product is the perfect fit
Draft natural replies that you can post in one click

Grow your business with Okara's Reddit Agent.

THE SOFTWARE BLOODBATH HAS A WINNER? (P2)

Alright, this is the part that most retail investors miss.

And it is arguably the single most important thing to understand about this company.

It's called the Ontology.

I know. The name sounds like a rejected Marvel villain, but stay with me because this is the whole ballgame.

Most companies trying to "do AI" take one of two approaches:

They throw an LLM at a pile of documents and pray it produces useful answers.
They build chatbots that can answer questions but can't actually do anything.

Both approaches have the same fatal flaw.

Here's the thing nobody talks about: ChatGPT was not trained on your business.

It doesn't know your processes, your supply chain, your warehouse logic, or your customer relationships.

Giving it access to your data is like handing someone a phone book and asking them to run your company.

Data alone is useless without context.

Palantir's approach is completely different.

The Ontology doesn't just give AI your data. It gives AI the context of how your entire business actually operates.

Think of it this way.

Say you run a manufacturing business.

You've got plants, warehouses, customers, shipments, all with complex interconnected relationships.

The Ontology models all of that into a live digital twin of your operation, built on three pillars:

The first pillar is data from virtually every source imaginable.

Palantir offers hundreds of out-of-the-box connectors that plug into ERPs, CRMs, homegrown systems, legacy databases, Snowflake, Databricks, and BigQuery.

The data can live on-prem, in the cloud, or at the edge; they connect to it all.

All of it gets unified into a single picture of the real world, actual flights in the air, patients in hospitals, shipments on the move, warehouses, machines, and work orders, not just scattered data points.

The second pillar is logic, which is where things get interesting.

The Ontology doesn’t just store data. It encodes how to think about that data.

How does a warehouse actually operate? That logic might live in an Excel sheet full of rules, a machine-learning model, a forecast, or a third-party optimizer.

It doesn’t matter where it lives; Palantir integrates it and maps it into the semantic object.

So the AI doesn’t just know your warehouse exists, it understands how your warehouse works.

The third pillar is actions, and this is what separates Palantir from chatbots and dashboards .

The Ontology doesn’t just describe the world. It lets you change it.

Creating a stock transfer order in SAP, triggering something on a factory floor, updating a financial system, firing a webhook, all of these become modeled actions that the AI can execute directly.

Layered across everything is security.

Every piece of data, logic, and action is governed by granular access controls.

Different teams see different slices of reality.

When AI agents are built on top, they inherit those permissions automatically, meaning the AI can only access and act on what it’s authorized to touch.

When you bring all of that together, data, logic, actions, and security, you get something nobody else has: a live operating system for your entire business.

It’s a lot, I know.

Quick side note for AI PRO members: there’s a private section later where I share whether this is actually in my portfolio or just under the microscope.

If you want the actionable part, feel free to scroll down and jump straight to that. For everyone still with me, here’s why this matters.

Source: @TradexWhisperer

Right now, most enterprise AI is what I'd call the "swivel chair" problem.

An analyst stares at a dashboard, gets an insight, then swivel-chairs over to another tool to figure out how to actually act on it.

Then they log into a third system to make it real. It's like having a brain that can think but no hands that can move.

The Ontology eliminates the swivel chair entirely.

Because the AI has access to the data, the logic, and the actions, it can reason about a problem and orchestrate the complex operations on the backend to actually solve it.

All in one system.

And because it’s all unified, you can generate a custom SDK (a software toolkit developers use to build applications) of your entire ontology, essentially an SDK of your business and its processes.

You can plug it into mobile apps on the plant floor, custom React apps for customer interactions, backend integrations, and agentic workflows inside your existing tools.

The Ontology becomes the API for your entire operation.

Here's the key insight: You cannot build a useful enterprise AI agent without an Ontology-like system.

An LLM by itself is just a language prediction engine.

It doesn't know that "Order #45721" is connected to "Customer ABC", who's connected to "Warehouse Delta", which has a maintenance schedule that conflicts with a delivery deadline.

The Ontology provides the context, not just the data, the operational grounding that makes AI actually useful in mission-critical environments.

Palantir spent 20 years building this system.

That head start is essentially impossible to replicate in a competitive timeframe.

So after all of that, here’s the question that actually matters:

Is Palantir a generational winner or a generational valuation trap?

Upgrade to PRO to access: