We’ve sent a lot of earnings reviews this season that can all be found here.

Oracle provides a slew of software and hardware tools for on-premise and cloud environments. It has 3 main segments which are closely integrated. 

Oracle Cloud Infrastructure (OCI) is its fully managed business for infrastructure services (virtual machines, storage, managed high-performance compute data centers etc.). This segment also includes platform services to build apps in its safe, controlled environment (serverless and container-based).

Strategic Software As A Service (SaaS) includes Oracle NetSuite. This is a set of applications for enterprise resource planning (ERP), customer relationship management (CRM), human capital management (HCM), e-commerce and more. It’s hard at work on launching more industry-specific software apps across areas like healthcare. It also has a more customizable, feature-rich product that is similar to NetSuite called Oracle Fusion. This one is geared towards larger customers.

The last segment is its broad range of database products including both relational and document-oriented offerings across structured, semi-structured and unstructured data. Creating valuable apps from GenAI infrastructure requires great models and rich, organized information access to properly season those models. That’s where its database capabilities come into play. 

Oracle closely integrates with the 3 big hyperscalers, allowing its databases to run anywhere. This also means that customers can migrate their on-premise databases to the cloud via OCI or through any of these hyperscalers, diminishing user friction. Oracle believes that this data cloud interoperability provides innate data transfer cost advantages. Cost is estimated to be “several times cheaper” for model training than any competitive product, according to leadership. 

Oracle has re-emerged as a digital infrastructure titan. While the company did take longer to roll out its high-performance compute product suite, it has since achieved fantastic traction.

Last month, Oracle announced they would raise $50 billion between debt and equity financing this year. They already issued $30 billion in bonds with great demand, but have yet to offer the planned $20 billion in equity (4.4% of the current market cap).

“This is all made possible by Oracle's transition from a predominantly seasonal license business into a highly predictable recurring revenue cloud business.” – Oracle CEO Clay Magouyrk

Oracle trades for 20x forward EPS (no cash flow right now). EPS is expected to grow by 23% this year and by 7% next year.

Oracle’s Thoughts on the Death of SaaS:

Leadership spent a lot of time talking through negative Software as a Service (SaaS) sentiment, and why they think it's sharply overblown. Like everyone else, they're confident that AI technology will profoundly reshape the way we all conduct work and deliver products. They just also think they're building the tools and the overall ecosystem to secure a large piece of that future landscape. Here we’ll discuss why they’re so bullish on the future.

They're using the technology to greatly accelerate their pace of application creation, while also embedding roughly 1,000 agents throughout existing products. They're offering customers tools to customize their own agents, without any upcharges. And they're doing all of this while offering customers a way to unify their database and application needs under one roof, driving interoperability, lower costs, and simplicity. Many of these customers are already parking their data in an Oracle database public cloud. They don't want more vendors, they want fewer. If Oracle is able to deliver impactful product innovation that keeps pace with all the change and disruption taking place, they should be able to continue providing unique value to their customers. And thanks to all of this first and third party data that they have at their disposal (not just Wikipedia, Reddit and other commoditized public data sources), they have a good chance of holding their ground.

Leadership spoke about conversations with customers not coming remotely close to mirroring the sentiment seen on Wall Street at the moment. There are no customers even contemplating ripping out any existing Oracle applications in favor of vibe-coded alternatives. The mountains of proprietary data, decades of compliance experience, highly complementary data and infrastructure services, and brisk AI innovation (that they don't have to do themselves) all continue to help Oracle drive high retention. AI is pushing customers to ask Oracle how they can embed more of this technology into the products they know and love. It is not tempting them to rip and replace with untested, unsupported, and unproven alternatives, conversationally created in 48 hours. Shocker (I say very sarcastically).

AI Halo Effect?

The dynamic discussed above is creating a platform-wide halo effect, where AI fosters incremental demand across its other product categories in a few ways. The need to embrace this new technology is pushing customers to migrate and modernize their existing data and application infrastructure, which helps Oracle’s products. Already housing so much of a customer's data is making Oracle an easy and convenient choice for running other AI-related workloads and using its infrastructure and data tools. All of the AI model training and inference taking place in its ecosystem is allowing rapid implementation of AI-driven upgrades to its applications. Examples of this include Fusion’s AI Agent Studio, which is being offered to customers as a free upgrade, permitting them to customize their own agents within the familiar Fusion interface. And its reliable infrastructure-related services are helping it win trust and business across its data and application pillars as well. 

As we already discussed, customers want fewer vendors, and Oracle offers great products across all three of these highly complementary buckets. Finally, because Oracle thinks its infrastructure build-outs are cheaper than competitors, it sees more budget being unlocked for application purchases, including its own. Simply put, AI helps everywhere. And because Oracle does so many things, it’s confident in readily using all of this AI advancement to benefit its customers and its own business in numerous ways. 

As a key aside, it's this overarching platform and the valuable products forming it that provide Oracle with confidence in winning a respectable piece of the AI agent opportunity that so many are vying for. 

Multi-Cloud Database:

The decision to allow Oracle Database Services to run in every public cloud, rather than just OCI, has proven to be a great one. It has unlocked a boatload of demand that drove 531% Y/Y growth for the quarter. A lot of this is driven simply by years of building interest in this product being offered through key partners like Azure, AWS, and GCP. They're now actually integrating across regions around the globe, and as that happens, demand predictably follows. Global region coverage now extends for all partner clouds, with 33 Azure regions live, 14 Google Cloud regions live, and 8 AWS regions live (with plans to grow that to 22 by the end of next quarter).

AI Infrastructure:

AI infrastructure revenue grew 243% Y/Y, driven by both GPU and CPU-related demand. Supply remains scarce for both product types.

Leadership spent a lot of time talking about how they're getting more efficient in thinking about and spending capex as they scale their data center footprint. They reviewed how Oracle's modular, Lego-like design allows them to standardize construction across any size of data center, in a way they view as more efficient than any other competitor. This efficiency is also why their GPU-based data center capacity laid this quarter remained two points above their long-term 30% GPM guidance. And CPU-based capacity continues to have margins in excess of that level.

They've reduced time from rack delivery to revenue generation by 60% over the last few quarters, and are also enjoying growing fixed cost leverage associated with growing into their quadrupled rack manufacturing capacity Y/Y. Additionally, the company is embracing a "bring your own hardware" form of data center build outs that shrinks the negative cash flow associated with expansion, while shifting to some upfront customer payments to accomplish the same thing. A key theme throughout the prepared remarks was the relationship between capex and negative cash flow untethering, as they get creative in how they structure these gigantic relationships. Contracts using this structure represented $29B in new business signed this quarter as a strong sign that customers are entirely fine with the changes. 

Next, Oracle leadership discussed 90% of their data center contracts being executed on, or ahead of schedule this quarter. In a world frantically trying to find supply and dealing with bottlenecks everywhere you look, that is strong execution. Alongside the cost advantages that leadership discusses every quarter, this continues to help Oracle gain the confidence of its partners in choosing OCI.

TikTok:

As broadly reported, Oracle now owns a 15% stake in the U.S. TikTok business. This won't impact revenue going forward, but they will record non-operating income based on that business's results and their proportional interest starting next quarter.

More Notes:

This was a good quarter. They're clearly capitalizing on the ongoing AI infrastructure tailwind, and their data and application products are also growing very nicely. Margins are likely going to feel significant pressure in the coming years as their GPU-based infrastructure business explodes at a 30% GPM, but they're confident in delivering profitable growth and exceeding previously communicated expectations. They're probably going to continue to do very well as long as this infrastructure cycle continues to zoom, and signs of that happening remain mostly positive at the moment.  Still, this is not a name I'm interested in owning. It’s just still nice to hear from them, as their large business and product suite spans AI data, apps and infrastructure. 

I think it's a great one to track for helping gauge the health of the AI cycle and how willing capital markets are to fund massive losses from AI darlings (its largest infrastructure customers). Their balance sheet is aggressively leveraged, their cash flows are deeply negative, their OpenAI reliance is significant and they're raising a lot of money right now. They would arguably be one of the first cracks to form as this super cycle finally begins to slow down.  It will be interesting to see whether or not their data and application product categories can accelerate quickly enough to offset that outcome. 

They look less resilient to me than other names in the overall AI infrastructure and compute category like Broadcom, Taiwan Semi and Nvidia. If I wanted exposure to this overall theme, I’d go with one of those companies instead.

My Rubrik Deep Dive, which covers the ins and outs of the business in detail, can be found here.

Business Model & an Accounting Note:

Rubrik has leaned heavily into a cloud-based subscription ARR-based model. It has proactively pushed customers from its licensing and maintenance revenue structure to annualized cloud subscriptions. As a result, cloud ARR is rising steadily as a portion of total. This form of revenue recognition is more visible, less lumpy and entails easier cross-selling to other products. It no longer relies on hardware sales or refresh cycles, which others are rapidly trying to get away from.

This transition includes Rubrik Security Cloud (RSC) credits to expedite cloud-based adoption. That’s leading to accelerated revenue recognition as it’s able to record allocated licensing reserves as revenue faster than it would have been able to. This is called material rights revenue. That has been a revenue growth tailwind this past year and will shift to a headwind during the new fiscal period.

Margins greatly benefited from the sharp demand outperformance. Q4 2025 was an abnormally strong period for cash flow generation. Annual free cash flow rose by more than 10% Y/Y from a small base.

RBRK trades for 38x forward FCF guidance. FCF is expected to grow by 36% this year and by 48% the following year. I would expect FY 2028 (not this year but the following year for them) estimates to rise following this strong report. The company has not been profitable for very long, so the most valuable valuation chart is using sales.

Cyber Resilience – Data:

RBRK was unsurprisingly asked about the threat of AI displacing their core data resilience businesses. They are not concerned. They think their increasingly large scale and experience gives them a base of proprietary data and edge-case know-how that vibe coding cannot emulate. They think their positioning in infrastructure software, rather than as a user interface wrapper, is quite defensible, and their increasingly broad presence across data, software partners, identity, and AI enablement should drive cross-selling. That, in turn, will create an acceleration in data collection and an ability to use that asset to stay ahead of the pack.

It's the same winning formula we're discussing for the other software names in the coverage network that I view as safe from this intimidating change. Proprietary data and an ability to enrich a broad range of specific use cases is far more defensible than tying various applications together with a pretty workflow that improves productivity. To me, that is what vibe coding can existentially threaten. 

Next, Rubrik’s data resilience offering is the final line of defense for a company recovering normal operations amid chaos and potentially destructive hacks. In this realm, you cannot accept any degree of hallucination rates prevalent among all AI models. AI is not capable of supplanting RBRK, in my opinion. Instead, it will exponentially enrich the work it’s doing and bolster the amount of information flowing through customer databases and applications. That should naturally support its core immutable (not modifiable) data backup and recovery business… it’s what they believe… and it’s what I believe. 

As leadership reminded us, this is why they are pursuing AI enablement with Rubrik Agent Cloud (RAC). They know being the company that enables large customers to lean into this technological wave with aggression and security will yield considerable cross-selling opportunities, expanding lifetime value, and an elongating runway. 

Next, they were rightfully intentional to point out that their business model does not rely on seat-based growth. We haven't seen seat-based headwinds preventing companies from meeting or exceeding guidance yet, but there are large layoffs happening across public markets. That makes a data volume-based monetization engine like this one comforting. 

RBRK's 90%+ competitive win rate against data protection vendors is concrete evidence of their product suite standing out vs. the crowded pack. The only deals they're losing are the ones they're not participating in, and the company's hard at work on making sure that happens less and less frequently. 

Finally, leadership views this category as still very early, with so much legacy displacement and overall growth left to enjoy.

Cyber Resilience – Identity:

This product continues to rapidly ramp faster than any other release in this company's history. In just three quarters, they've gone from zero to 900 customers, and more than doubled Q/Q. During the quarter, RBRK deepened their Microsoft relationship by integrating their identity resilience product with M365. This led to 50% of M365-based RBRK bookings, including their nascent identity solution. They also added identity-based support during the quarter for Okta, which is already notching material traction just a couple months in. This gives them integrations with that popular directory alongside Microsoft’s Active Directory and Entra ID, which they believe is the only identity recovery platform to effectively span all three. 

They talked about a couple large identity customer wins coming thanks to the millions in cost that Rubrik allows these firms to eliminate when the inevitable breach occurs thanks to faster and easier operational normalization.

Rubrik Agent Cloud (RAC):

This product entered general availability during the quarter. RAC's ability to offer an overarching view of agentic activity while featuring deep and granular permission controls is proving to be a popular combination. These are really the two RAC use cases that are building the most traction early on, but they expect a compelling Agent Rewind product to quickly gain steam in the coming quarters. As a reminder, this allows companies to recover the most recent normal state of an agent's operations, while identifying what went wrong to eliminate re-infection. That means significantly less disruption, just like with its data and identity cyber resilience products. It also means companies can feel a lot more free to liberally test and adopt cutting-edge AI, without risking catastrophe. As leadership correctly said, agents provide 100x more productivity and 100x more risk. Proper guardrails are vital to make sure the damage from constant hallucinations is manageable.

They're in the proof of concept (POC) phase with a few large customers, and if all goes well, things should start to ramp from there. Leadership is being very cautious in how they talk about the potential maturation of this product, but you can tell they're excited… and they should be.

There are many companies going after the same opportunity, but RBRK is confident that they'll win their fair share. They think they've built an intent-based design for RAC that will stand out vs. what they view as rules-based competitors that have configurations that are too rigid and static. The team is excited about Predibase’s (recent acquisition to provide foundation for RAC), model customization platform and training/inference optimization. They’re confident these tools will be highly successful complements to their existing platform that will help differentiate RAC vs. substitutes. I agree. Allowing customers to make sure their model training and inference is more efficient with you than other vendors is a great way to ensure they're housing more data, and becoming more reliant on your core business. Many people think of AI enablement as a random adjacency for RBRK, but to me, it's so related to and supportive of what they already do; I think things like this make that clear. 

And finally, they also believe the ability to interoperably tie these products to necessary cyber insurance and resilience tools will be a strong selling point. That's likely why ServiceNow is partnering with Cohesity (an RBRK competitor) to make sure they have this integrated offering as part of their own agent monitoring and orchestration platform. Like enterprise software, I think this will be a massive and fragmented opportunity, with many winners taking a small piece of the gigantic total addressable market and enjoying strong growth in the process. 

More Notes:

Strong quarter. I think this company continues to show how durable its growth engine is amid all of the disruptive AI-based change happening throughout software. I think they're poised to greatly outperform their initial annual guidance and should deliver close to 30% Y/Y ARR growth at an increasingly large scale. They'll do that while briskly improving margins, successfully expanding into new product categories, and proving to the world that their value proposition cannot be vibe-coded away. The ramp for their identity business is nothing short of impressive, and I think their AI enablement arm will look very similar over the coming quarters and years. That gives them two highly compelling growth vectors to join its core data cyber resilience niche, which continues to have so much opportunity remaining. I continue to be a confident shareholder, and would look to add to my stake if Mr. Market keeps punishing this name alongside the entire cohort.