Cloud Wars Live with Bob Evans

In this episode of Cloud Wars Live, Bob Evans sits down with Tirthankar Lahiri, Senior Vice President for Mission-Critical Data and AI Engines. Lahiri explains how agentic AI is transforming enterprise applications from simple question-answer systems into action-driven platforms that can reason, remember, and securely execute tasks. He details Oracle’s strategy around unified agent memory, private agent factories, deep data security, and open development standards, all designed to help customers build scalable, secure, and flexible AI systems without added cost.

AI Built Securely

The Big Themes:

Agentic AI Becomes Action-Oriented: Tirthankar Lahiri explains that agentic AI represents the next major step beyond generative AI. While generative AI focused largely on answering questions and producing content, agentic AI is designed to take action. It allows businesses to build systems that can reason, decide, and execute tasks autonomously. Oracle sees this as the future of application development, where AI becomes embedded into workflows rather than functioning as a standalone tool.

Oracle Builds AI Directly Into the Database: Rather than forcing customers to move data across multiple isolated systems, Oracle’s approach is to bring AI directly to the data. Lahiri argues that data is the “ground truth” and moving it creates technical debt, silos, inefficiency, and security vulnerabilities. Oracle’s converged database architecture supports multiple data types, including relational, graph, spatial, and vector, inside one unified environment. This eliminates the need for separate repositories and allows AI agents to access all relevant context without fragmentation.

Deep Data Security Protects Against AI Risks: Lahiri strongly emphasizes that traditional application-layer security is no longer enough in the age of AI. Since AI can generate SQL and potentially bypass interface restrictions through prompt injection, businesses must secure data directly at the source. Oracle calls this “deep data security.” He uses the analogy of protecting valuables in a safe bolted to the floor rather than simply locking the front gate. Even if someone gets inside the house, the valuables remain protected. Similarly, Oracle enforces security policies at the database level, ensuring agents can only access data users are authorized to see.

The Big Quote: "You need to secure data. Need to lock your valuables into the safe deep inside the house."

More from Tirthankar Lahiri and Oracle:

Connect with Lahiri on LinkedIn or learn more about Oracle AI Database.

Visit Cloud Wars for more.

In today’s Cloud Wars Minute, I break down Google Cloud’s $750M partner ecosystem investment and its aggressive push to lead the agentic AI transformation race against Microsoft, AWS, and Oracle.

Highlights

00:03 — Last week in Las Vegas, we saw Google Cloud roll out an incredibly broad, deep set of agentic AI technologies and solutions. They’re making a massive push on this. Google Cloud is the number one player in the Cloud Wars Top 10, and they’re looking to really take the big lead in agentic AI transformations for their customers.

00:24 — Now, in parallel with that, we also saw Google Cloud — in addition to all what they’re doing on the tech track — on the go-to-market side. They have put together what they’re calling a $750 million fund for their ecosystem and partners to help them drive AI transformations for customers, right — to help them get better training, to be more technically proficient, to help deploy engineers, and more.

00:53 — So I’m wondering now, in light of this, both the tech splash that Google Cloud made at Next but also now with this push for their partner ecosystem, will Microsoft and AWS match this massive, I would say unprecedented, investment in the ecosystem?

02:26 — You’ve got, not so long ago, I think you could look at what we called broadly the hyperscalers, and it was a little bit hard to differentiate, right? They were just sort of known as this one blob. I think the companies Microsoft, AWS, and Oracle are differentiating themselves now to a vast degree.

02:47 — I think what Google Cloud is trying to do here is say, “Hey, we’re going to do all the stuff we did for you before, but we are going to focus more intensely than any company on Earth on the agentic AI revolution, to provide not just the technology, but the skills that you need from the partner ecosystem."

03:37 — And then it's putting the money behind this to ensure that those partners have the capabilities to do it — and to do it very quickly — because this is a race to see who is going to get that first-mover status. So far, I believe Google Cloud is in that for agentic AI transformations for customers across the globe and across industries.

Visit Cloud Wars for more.

In today's Cloud Wars Minute, I break down how SAP is outperforming Oracle, Salesforce, Workday, and Microsoft in the applications race.

Highlights

00:03 — We saw some first-quarter numbers from SAP late last week, and I think it's fascinating to see how, in the light of these very, very strong numbers from SAP, we still seem to hear about this idea that these doomsayers are saying that AI is going to destroy the enterprise apps business. That's certainly not happening.

00:33 — The three big numbers here: cloud revenue up 27%, almost $7 billion. Within that, its Cloud ERP Suite was up 30% to $6.1 billion, and looking out at contracted business not yet recognized as revenue, it calls it current cloud backlog, up 25% to well over $25 billion.

00:57 — That's pretty healthy-looking business for one that is, you know, doomed to the apocalypse and Armageddon any day now, according to some of these wizards of smart who believe that AI is going to come in and just decimate the enterprise apps business.

01:52 — The only thing I can think of that gives them this idea that the whole industry is heading for Armageddon here, they must believe that companies like SAP are unprepared for or unable to participate in the AI revolution, but that notion ignores where the data really is.

02:28 — Agentic AI is the future. Agents need data to run, and who has more and better business data than SAP? So this just confounds me.

02:43 — We continue to see SAP outperform Oracle's applications business, Workday, Salesforce, and also the apps business part of Microsoft Dynamics 365. I think this is a very strong quarter by SAP, and the future here is very bright.

Visit Cloud Wars for more.

ogle Cloud Next, Cloud Wars CEO and Founder Bob Evans sits down with Karthik Narain, Chief Product and Business Officer, Google Cloud, to discuss how AI is fundamentally changing enterprise expectations. Narain explains why customers no longer judge technology providers by licenses sold or cloud consumption, but by measurable business outcomes. From forward-deployed engineers to agentic workflows and the evolving role of product design, he outlines how Google Cloud is rethinking engagement, product development, and enterprise transformation in what he calls the third major wave of technology innovation.

AI Demands Outcomes

The Big Themes:

Outcomes Replace Consumption Metrics: Narain explains that enterprises are no longer measuring technology providers by how many licenses they sell or how much cloud consumption occurs. Instead, success is now judged by outcomes delivered. Customers expect providers like Google Cloud to share equal responsibility for business results, not just provide tools and leave execution to the customer. This represents a major shift from prior eras where businesses viewed themselves as the sole owners of converting technology into value. AI’s speed and sophistication have raised expectations dramatically.

The Third Technology Wave: Narain frames today’s AI era as the third major wave of enterprise technology over the past 60 years. The first wave from mainframes through ERP, focused on codifying business processes into repeatable systems. The second wave centered on delivery model innovation, moving software into SaaS and cloud environments. The third wave is fundamentally different because the technology itself learns and evolves. Rather than giving software fixed instructions, enterprises must feed it data, context, and reasoning. This changes how software is designed and deployed.

Every Feature Must Become a Skill: Products must now be designed for both humans and AI agents. Narain explains that every feature inside enterprise software needs to be exposed as a “skill” that agents can activate directly. This means software can no longer assume a human user is the only operator. Agents must be able to trigger workflows, execute tasks, and coordinate processes independently. This changes how products are structured from the ground up.

The Big Quote: “The application’s user interface is no longer clicks and drop-downs. It is going to be prompts and agentic workflows."

Visit Cloud Wars for more.

In this Cloud Wars conversation, Bob Evans sits down with Shub Bhowmick, CEO and Founder of Tredence, alongside Yasmeen Ahmad from Google Cloud to explore how enterprises are moving from AI applications to AI agents. Their discussion focuses on what it takes to turn intelligence into action — covering data foundations, semantic layers, agentic architectures, and the operational shifts required for businesses to scale AI successfully.

Turning AI Into Action

The Big Themes:

AI Agents Redefine Applications: Traditional AI apps assist by querying data, generating recommendations, and supporting limited workflows. AI agents, however, represent a much deeper operational shift. As Ahmad explains, agents are multi-step reasoning engines that can access multiple systems, coordinate actions, and execute entire business processes autonomously. Instead of simply helping humans decide, they can perform work across ERP systems, supply chains, and customer interactions. This changes the role of the database itself — from a storage and query engine into a reasoning engine with vector search, graph RAG, and semantic understanding. Examples like Home Depot and Danfoss show how this creates massive efficiency gains

Why Questions Require Agentic Intelligence: Shub Bhowmick draws a critical distinction between “what” questions and “why” questions. A conversational BI system can answer what happened — such as how much sales dropped — but a “why” question demands deeper reasoning. Why did sales decline? Was it pricing pressure, competitor behavior, inventory constraints, or macroeconomic events? These problems require hypothesis-driven exploration. Tredence addresses this through business semantic layers, knowledge graphs, and hypothesis banks that support open-ended reasoning.

Closed Systems Create Long-Term Risk: Bhowmick warns against enterprises rushing toward closed, inflexible platforms simply because they promise faster short-term value. While packaged solutions may accelerate deployment, they often restrict ownership, adaptability, and future innovation. In contrast, open architectures built with hyperscalers like Google Cloud allow customers to own the IP, customize solutions, and evolve as the market changes.

The Big Quote: “Gone are the days when these migrations used to take 12 to 18 months. Nowadays, you have to complete these migrations in less than three to four months.”

More from Tredence and Google Cloud:

Learn about the partnership between Tredence and Google Cloud and AI agents on Gemini Enterprise.

Visit Cloud Wars for more.