Cloud Wars Live with Bob Evans

Highlights

00:02 — Workday last week reported its Q2 results. They came in pretty strong, and I think if we look at the tenor of some of the comments that co-founder and CEO Aneel Bhusri made on the Q2 earnings call last week, we'll see that he's generally pretty low-key. But he's quite bullish now about what's going on with the early momentum Workday is gaining from its AI products.

00:49 — First, quick overview of some key Q2 numbers. Subscription revenue was up 13.9% to $2.47 billion. Its current RPO was up 14.2% to just over $9 billion. And for all of fiscal '27, which will end January 31, its guidance now says that the company expects to grow 13% to about $9.945 billion.

01:54 — AI tools drove more than $100 million in annual contract value in Q2. Workday is very excited about that and believes that that shows good evidence of the uptake from customers and the rising demand for these AI solutions. Also, the AI run rate for AI SKUs, as Workday calls it, across the company is now almost $600 million, and it said that's up 200%.

02:55 — And then what I'm calling the power of enterprise context. Bhusri said said, "We're an enterprise context company." And he said that context — who reports to whom, who pays the bills, where different people are organized, and so on — all those little sort of unstructured things that go on within corporations. Workday has it.

03:35 — So, it has been six months now since Aneel Bhusri returned to Workday as CEO. He felt that with the rise of AI, the AI revolution, that it was time Workday needed to have a deep technologist in the CEO spot, leading the company through this tumultuous change. So, he's increased the pace of technology innovation. He has got the growth rate moving back upward.

Visit Cloud Wars for more.

In today's Cloud Wars Minute, I break down Microsoft’s evolving AI consumption strategy and how it compares with Salesforce, UiPath, and other major industry players.

Highlights

00:19 — Microsoft just announced the GitHub Copilot harness for Copilot Studio — a really cool new capability that enables an agentic reasoning loop.

00:33 — As part of this release, it said that on September 1, it's going to start charging for consumption on the GitHub Copilot harness specifically. This is an interesting move for Microsoft because 90% of Fortune 500 organizations have actually given developers a way to be able to explore their curiosity and consume this technology

01:12 — To compare and contrast in this space, Microsoft on the GitHub Copilot harness is giving $0, if you use the Copilot Studio standard harness, there's actually free allocation for developers to go play with this. Now, if we compare this to other players in the industry, such as Salesforce, which Agentforce provides 200,000 Flex Credits to be able to come in.

02:05 — What I would tell you is that I'm expecting that this is going to not necessarily be something that will be able to continue in a market where there's a lot of demand. Just to give you perspective, when GitHub Copilot was doing this before, it took about seven weeks for them to realize that there was a problem with this, and they corrected it.

02:50 — This is lining up to Satya Nadella back on July 29th, saying that it's evolving its business model to go beyond per-seat licenses and are leaning into consumption. So, we will continue to see these types of things, and I expect to see other growing pains like this for all vendors as consumption starts to become the new metric.

Visit Cloud Wars for more.

In today's Cloud Wars Minute, I look at how Google Cloud is expanding the Gemini ecosystem into the automotive world through its partnership with Mahindra.

Highlights

00:03 — Google Cloud is partnering with the Indian automobile manufacturer Mahindra to get Gemini Enterprise installed directly into its vehicles. That's another interesting example that we're seeing here of AI leaping out of the digital space and becoming something much more physical.

00:23 — In this new development, the automotive agent in Gemini Enterprise is being integrated into MAIA, Mahindra's Artificial Intelligence Architecture platform. Beyond being a voice assistant, Gemini can understand natural language requests and actually interact with a car and its functions.

00:41 — Sashi Sreedharan, Managing Director of Google Cloud India, explained that "by enhancing its in-vehicle assistant with an automotive agent built on Gemini Enterprise, Mahindra is creating personalized and helpful conversational interactions for navigation, search, media communication, and other vehicle controls that will help drivers navigate and explore rich features in a more natural and helpful way."

01:13 — What we're seeing here is pretty significant: AI operating at the edge, with this idea of the digital and physical working together, and the car becoming the AI endpoint. This is also a really important way for Google Cloud to expand the Gemini ecosystem beyond smartphones, PCs, and enterprise software.

01:37 — The real opportunity that I see here isn't necessarily about Gemini as an AI assistant, but where the tech is embedded; it's those all-important endpoints.

Visit Cloud Wars for more.

In today’s Cloud Wars Minute, I look at how AI monetization is moving beyond execution and into creation as agentic development becomes more costly and valuable.

Highlights

00:15 — One of the things that I want to talk about today is the meter moving upstream. A lot of this has to do with the fact that what we're starting to see is that the cost that is associated with the actual build process is starting to get really high. And the reason behind that is the concept of an agentic reasoning loop and the cost that's associated with the build process.

01:35 — And what you're starting to see is that the build is actually costing a lot more because of the fact that we're using these different AI models in an agentic reasoning loop to be able to go do the build process. An example of this would be Microsoft starting to charge for the creation of agents inside of Copilot Studio.

02:03 — This is not something that's new in the industry because of the fact that we see other vendors that have tried this across the board. An example is that is exactly what you see in GitHub Copilot and Claude Code. These are definitely going to be tools that are building and being able to execute, and we're seeing a massive amount of adoption in this space to be able to allow builders to be able to build.

02:49 — We definitely are going to see some movement in this space. We're probably going to see each of these different vendors want to compete on who can give the best experience and do it in a cost-effective way. It's going to be very interesting to see how this starts to play in the concept of discoverability of different products and of different vendors.

Visit Cloud Wars for more.

In today's Cloud Wars Minute, I analyze Salesforce’s latest results and why Dreamforce could be critical for proving its AI opportunity is here now.

Highlights

00:03 — Salesforce yesterday released its Q2 results, and I'd say that while there's some very good things in there, it's a little bit of a mixed bag relative to what some other companies in its category are doing. But let's take a look at this because I think the big, the big news here, aside from the numbers, is Salesforce announced a huge new partnership with Anthropic.

00:41 — Customers will be able to tap into the power of Anthropic reasoning with their Salesforce applications and agents and data. So I think that offers a lot of potential for the future. But some of the key numbers here: we saw Salesforce revenue go up 11% for Q2, ended July 31. That's $11.3 billion relative to Q1.

01:30 — Its AI and data products—starting to break those out—said they now have an annualized revenue run rate of about $3.85 billion, almost $3.9 billion. It said that's up 210%, and then for Agentforce, annualized run rate of almost $1.5 billion, it said that's up 240%, and this, that.

02:43 — Combine that with the incredible presence of Salesforce applications and agents all over the world, all sorts of businesses. Bring those together. I'm sure for Anthropic, this is great too, as they aspire to become an enterprise powerhouse. The Salesforce partnership and connection here gives them entrée to some huge customers and huge opportunities.

04:21 — It's got to step up the urgency on that, and I think the deal with Anthropic is going to be just the medicine that's needed here for Salesforce. Got a detailed article that'll go into this on CloudWars.com, and we'll be following up more next week with the analysis of what Marc Benioff and others talked about on the earnings call.

Visit Cloud Wars for more.

In today's Cloud Wars AI Minute, I explore why falling token prices aren't necessarily making AI transactions cheaper.

Highlights

00:01 — So, one of the things that I want to talk about today is going to be the token economy. And what we're actually seeing is that tokens are about 12 times cheaper than they've been before.

01:30 — But what we're seeing is the average cost of a transaction is actually going up, and that is because of the fact that while token prices are coming down, the amount of tokens it takes to process something is actually going up. We're actually seeing a 520 percent increase in the amount of tokens that you're going to need to have per unit of work.

01:56 — And the reason behind this has been this agentic reasoning loop and the reasoning models that are starting to go and process. As a matter of fact, you're actually seeing that most recently, we've even seen that chat-based things and chat conversations that have happened between humans and agents have actually now, for the first time, started to fall off.

02:30 — So, if this industry trend continues, what you're going to see is that we're going to continue to see this drop. But what we want to see is we want to start also seeing some sort of normalization and drop happening on the average amount of tokens spent on a per-transaction basis.

Visit Cloud Wars for more.

In today's Cloud Wars Minute, I explain why Workday believes AI security is becoming an ecosystem problem that no single vendor can solve independently.

Highlights

00:06 — Workday has become the latest organization to join the Open Secure AI Alliance. It's a well-supported initiative that aims to improve the security of AI systems through open-source tools, research, and knowledge sharing. Workday's joining this alliance, and it's in very good company alongside companies like NVIDIA, Cloudflare, Microsoft, and IBM.

00:31 — I want to read a short passage from Workday's joining announcement to give you a sense of why the company chose to join the alliance. So, from Workday: “Delivering safe, trusted enterprise AI requires multiple layers of protection across the entire application stack, from identity management and access permissions to continuous guardrails, monitoring, and governance.”

00:53 — “Our commitment to continuously build our internal security engineering programs to meet these high standards is why Workday is joining the Open Secure AI Alliance.” Enterprise AI is shifting quickly towards more open stacks and more ambitious agents. Initiatives like the Open Secure AI Alliance are incredibly important.

01:22 — Workday and the other members of the alliance recognize that AI security is quickly becoming an ecosystem problem, not just something that one vendor can solve independently. And that's why, for example, Workday specifically said that it has historically worked on security internally for its own platform, but now is sharing what it's learned with the wider community.

01:50 — As AI becomes more open, distributed, and agentic, securing models isn't going to be enough. The challenge now, and that's one that Workday and the other members of the alliance are trying to address, is securing everything around these models.

Visit Cloud Wars for more.

In today's Cloud Wars Minute, I examine why the emerging Google Cloud and Oracle alliance could reshape how enterprises accelerate their AI transformations.

Highlights

00:01 — Interesting to see some new and different combinations of partners moving forward here in the AI Revolution, because the pace at which customers have to move, their escalating business, competitive pressures, their new technology requirements, and just the urgency they're all feeling to innovate, move into their AI transformations as quickly as possible, is leading to some unexpected combinations of companies and new sorts of enhanced partnerships.

00:33 — So, I think a real power pair that is emerging are the number one and number two companies on the Cloud Wars Top 10, Google Cloud and Oracle, and I think they're coming together. Each has sort of put forward into the mix, not some peripheral products they have, but really, in each case, sort of the crown jewels for what's happening right now.

01:26 — Google Cloud for Q2 had revenue of almost $25 billion, up a remarkable 82%. Meanwhile, its backlog soared 390% to $514 billion. Oracle, for the quarter ended May 31 — and it'll have its new numbers out in a couple of weeks — but for the quarter ended May 31, its fiscal Q4, its cloud revenue was almost $10 billion, up 47%.

02:50 — So, really, each company has put forward the very best it has to offer in this new and extended partnership. And again, I think we're going to see more and more of these sorts of groundbreaking, unprecedented, or certainly unexpected partnerships going forward, because what customers want and need now is changing so rapidly.

03:47 — So, this is a great, great new step forward for customers. And again, I think we're seeing lots of different combinations among the Cloud Wars Top 10 companies. So, these partnerships, alliances, whatever you want to call them, they're more important than ever before.

Visit Cloud Wars for more.

00:03 — According to recent reports, Microsoft is set to reveal its Maia 300 AI accelerator as early as September this year. For context, Microsoft launched Maia 200 back in January with the direct aim of reducing the cost of AI inference.

00:27 — The Maia 300 is expected to support both AI training and inference, but the key story here is the scale of ambition behind it. Now, Microsoft is once again partnering with TSMC, Taiwan Semiconductor Manufacturing Company, who delivered the first iteration of Maia chips. However, this time, the scale is much larger. increasing from tens of thousands of Maia 200 accelerator units.

00:54 — It's increasing from tens of thousands of Maia 200 accelerator units to potentially hundreds of thousands, or even over a million Maia 300 chips. For me, this initiative is really about Azure. By investing in these chips, Microsoft isn't necessarily trying to compete with NVIDIA. Rather, it aims to ensure that Azure doesn't rely entirely on NVIDIA infrastructure.

01:14 — This move will give Microsoft greater control over the cost, supply, and performance of the AI infrastructure running on Azure. And beyond that, Microsoft CEO Satya Nadella has stated that building the company's own silicon to work alongside NVIDIA and AMD chips is a priority.

01:35 — The successful launch of Maia 200 demonstrated that Microsoft's custom silicon strategy can work, albeit on a relatively small scale. The hope now is that Maia 300 will enable Microsoft to scale this strategy into a genuine competitive advantage for Azure.

Visit Cloud Wars for more.