Boomi, IBM point to AI agent orchestration inflection point

Larry Dignan

Editor in Chief of Constellation Insights
Constellation Research

Larry Dignan is Editor in Chief of Constellation Insights at Constellation Research, where he leads editorial coverage focused on enterprise technology, digital transformation, and emerging trends shaping the future of business. He oversees research-driven news, analysis, interviews, and event coverage designed to help technology buyers and vendors navigate complex markets with clarity and context. ...

Read more

It's easy to build an AI agent and even scale them. It's much harder to orchestrate them to drive returns. AI agents need a manager pronto.

Those are the takeaways from Boomi CEO Steve Lucas and IBM's Bruno Aziza, Group VP of Data, BI and AI. Aziza also is a member of IBM Ventures' investment committee.

Both tech leaders spoke at Constellation Research's Connected Enterprise conference where AI agent sprawl and orchestration was a recurring theme.

Boomi calls BS on inflated AI agent numbers

Boomi CEO Steve Lucas said enterprises are moving to AI agents as they shift from deterministic processes to probabilistic processes, but be wary of inflated claims about how many are being deployed.

"We live in this world called reality," said Lucas. "We've deployed roughly 50,000 agents in our customers that are real with code and grounding for business processes. That's not millions and anyone that tells you millions is full of shit. It's not true."

Lucas added that AI agents need to be judged on returns and real business process impact. "We're seeing real ROI," said Lucas, who noted Boomi is deploying internally and with customers.

He added that there's a ground game to consider. "When we walk into a client today we start with a workshop. You can build a new agent with a chat interface, but do they have the right data and process?" asked Lucas.

Next up will be the orchestration of agents, which is often more of a vendor talking point. "How do I monitor for effectiveness and ROI? You have to determine the ROI and whether agents are overlapping all day," said Lucas.

Boomi has an agent builder that includes governance, but will extend the tool to monitor ROI and orchestrate multiple AI agents. "What we're going to be announcing next year is the multi agent orchestration extension. We're not building an agent control towers. We don't just watch AI agents work. You need to orchestrate multiple agents and figure out do you have the right model and what are the costs."

Lucas said 2026 will be about creating an AI activation layer and moving beyond automation. "It's not just about creating agents it's going to be about orchestrating," said Lucas.

• For AI agents to work, focus on business outcomes, ROI not technology
• Boomi acquires Thru, makes case for AI orchestration, automation platform
• Boomi World 2025: Agentstudio, AWS pact, 33,000 AI agents deployed
• Boomi launches AI Studio, makes its move on AI agent orchestration

IBM's Aziza: AI agents have no manager right now

Aziza said it's easy to create an agent--he has five or six GPTs to serve up answers to his kids.

He said:

"Building an agent of agents is really simple. The problem stuff that's going to get us in trouble is you're going to have multiple agents built across multiple platforms, and no single vendor will have the incentive to just help you manage and  orchestrate across multiple platforms. AI has no manager right now."

The state of play today is the following:

• Vendors are making it easy to build agents for developers as well as business users.
• Enterprises can choose to stay on one platform for agentic AI, but that's limiting.
• The challenge will be finding the vendor that's horizontal and "coming to help you orchestrate and operationalize these agents."

IBM TechXChange 2025: Big blue connects agentic AI, mainframe dots, partners with Anthropic

"The key word for us in the next 10 years is going to be multi-agent orchestration," said Aziza.

Constellation Research’s sneak peek at 2026

Larry Dignan

Editor in Chief of Constellation Insights
Constellation Research

Larry Dignan is Editor in Chief of Constellation Insights at Constellation Research, where he leads editorial coverage focused on enterprise technology, digital transformation, and emerging trends shaping the future of business. He oversees research-driven news, analysis, interviews, and event coverage designed to help technology buyers and vendors navigate complex markets with clarity and context. ...

Read more

Constellation Research analysts were split on whether we're in an AI bubble or not, recapped 2025 in AI agents and gave a hint of what's to come in 2026 around decisions, automation and exponential efficiency.

Here's a look at what Constellation Research analysts said about 2025 and 2026 at the opening panel of Connected Enterprise 2025.

Michael Ni

2025: The great divide between enterprises is not about digital. "It's actually a decisional divide and it will decide winners and losers," said Ni. "Winners and losers will really be about who can actually get to decision velocity while keeping compliance control."

2026: "You'll see a massive shift toward a decision centric architecture. Unified data foundations will need to extend to context and grounding. You'll see process and workflow automation incorporate decisions and governance."

Chirag Mehta

2025: "The biggest adoption barrier for AI is security, control and trust. It's not a lack of use cases. It's not a lack of money," said Mehta. "Enterprises want control. They want a specific outcome system and that's driving a lot of innovation and spend in cybersecurity, which was traditionally considered a cost center. Now people are looking at ROI for security and it's amazing."

2026: "Right now people are building AI agents, but they have no idea how to manage a life cycle of an agent, access and privileges. The innovation will be in AI agent identities."

Me

2025: "Agentic AI is the obvious trend and everyone and their mom wants to be a platform. And this game of musical chairs isn't going to go well."

I also noted that we need to start segmenting the AI market. AI infrastructure is bubblelike. Enterprise AI hasn't really started yet. It's time for nuance.

2026: "We're going to realize later in 2026 that all we did was scale mediocrity with the same LLMs and data. We're going to need to come up with ways to be creative and innovative."

Liz Miller

2025: "The challenge for marketing tends to be that you're always being sold a silver bullet," said Miller. "We've seen so many bubbles. The real trend is that marketing is actually using AI. We're really on the front line, but there's a gap between the expectation and the delivery."

2026: Marketing departments will start hiring people because they'll realize the limitations of agents.

Esteban Kolsky

2025: "The biggest trend right now is the commoditization of subsidized AI."

2026: "In the second half of 2026, we're going to start investing in what really matters. Private platforms with a distributed computing architecture.

Martin Schneider

2025: "Revenue ops will optimize using AI but there will be smaller models that are precise and can get better understanding of workflows," said Schneider.

2026: Revenue ops will continue to be revamped via AI agents.

Holger Mueller

2025: "There is a broad revival as everything is going agent with HCM." the disappearance of the divisional / Departmental HR people.

2026: "There will be frontline worker empowerment to complement the agents that are being built," said Mueller. He also said that AI agents will begin to replace divisional and departmental HR people. 

Ray Wang

2025: "There's a romantic notion that there will be an agent per persona. We're going to tell you that it's false. You're seeing automation push across the back office."

"We're seeing a manufacturing renaissance. It's supply chain, precision manufacturing, data centers, distribution and energy."

2026: "One of the biggest things we'll be talking about is the notion of exponential efficiency. You're seeing revenue per employee going from $100,000 to $1 million to $10 million," said Wang. "We're going to see the difference between winners and losers based on who adopted AI."

Microsoft Dynamics 365 expands sales AI agent roster

Larry Dignan

Editor in Chief of Constellation Insights
Constellation Research

Larry Dignan is Editor in Chief of Constellation Insights at Constellation Research, where he leads editorial coverage focused on enterprise technology, digital transformation, and emerging trends shaping the future of business. He oversees research-driven news, analysis, interviews, and event coverage designed to help technology buyers and vendors navigate complex markets with clarity and context. ...

Read more

Microsoft said its Dynamics 365 Sales Qualification Agent and Quality Evaluation Agent are generally available with its Sales Close Agent and Sales Research Agent in public preview.

The rollout follows a blistering pace for AI agent losses in recent days. Salesforce had its Agentforce heavy Dreamforce conference, Oracle layered AI agents throughout its Fusion applications, Google launched Gemini Enterprise as an agent orchestration platform and AWS said its Bedrock AgentCore is generally available.

You could say it's an agenteriffic time in enterprise technology.

For Microsoft the general availability of its agents come as the company is evolving Dynamics 365 from a system of record to a system of action. Dynamics 365's future will revolve around agentic business applications.

• Microsoft Build 2025: Agentic AI Meets Enterprise Data Decisioning
• Microsoft Keeps Pace in Multi-Agent AI Race
• Microsoft wants to be your agentic AI developer stack

Microsoft is starting its agent parade with sales functions. Sales Agents in Dynamics 365 Sales are mapped to "jobs to be done." These jobs are encompassed in various roles such as sales development reps, sales reps and sales leaders. Here's how it breaks down:

• Sales Qualification Agent is designed to assist a sales development rep who has a goal of generating pipeline. The agent will carry out tasks such as researching leads, customer outreach, qualifying leads and ultimately converting them. Sales Qualification Agent can engage hundreds of leads at once with personalized emails and follow-ups. It also identifies customer fit and stage and routes to seller after qualification.

• Sales Close Agent will assist sales reps trying to win deals and will cover handoffs, discovering use cases, proving value and negotiating and closing deals. Sales Close Agent can also autonomously engage customers, follow long-running deals and mitigate risks and provide insights to stakeholders.
• Sales Research Agent is for sales leaders on the hook for growing revenue and optimizing the business. Managing pipeline, team account and territory planning and sales ops are core tasks. Sales Research Agent automatically generates research plans and uncovers patterns.

These agents join various offerings covering customer intent, customer knowledge management and case management in Dynamics 365 Contact Center and Customer Service.

• Anthropic, Microsoft, OpenAI build out AI agent use cases
• Microsoft delivers strong Q4, Azure delivers $75 billion in annual revenue
• Microsoft 365 launches role-based Copilots

Oracle’s application strategy: Industries, AI, automation, integrated stack

Larry Dignan

Editor in Chief of Constellation Insights
Constellation Research

Larry Dignan is Editor in Chief of Constellation Insights at Constellation Research, where he leads editorial coverage focused on enterprise technology, digital transformation, and emerging trends shaping the future of business. He oversees research-driven news, analysis, interviews, and event coverage designed to help technology buyers and vendors navigate complex markets with clarity and context. ...

Read more

Oracle's Fusion applications business can't compete in buzz or growth rates compared to Oracle Cloud Infrastructure (OCI), but you'd be wise to pay attention to the software plan. Yes Virginia, Oracle is still a software company.

At the just concluded Oracle AI World, the buzz machine revolved around OCI as well as an AI platform and data lakehouse play. Fusion application news landed on Oracle AI World's last day. For what it's worth, Oracle is scaling up AI agents across its applications focused on various processes like accounts payable, payments, fulfillment, legal and financial planning and analysis.

In a week full of AI agent hubbub due to Salesforce's Agentforce barrage, you are forgiven for missing the Oracle AI agent news. But it's worth catching up.

• Mike Ni: Oracle AI World 2025: Shifting left AND right into Oracle’s suite spot
• Oracle CTO Larry Ellison on AI use cases for healthcare, agriculture, climate change
• Oracle AI World 2025: Autonomous AI Lakehouse, AI Data Platform launched

Steve Miranda, Executive Vice President of Oracle Applications Development, said at the company's investor day at Oracle AI World:

"What I talked about at this conference last year was that we would have 100 AI agents in Fusion. We actually have 600 AI agents, 400-plus in Fusion, 200-plus in the industry verticals. The 600 are agents that we've built and that does not count agents that our customers are building through the agent studio or what our partners have built."

Oracle also launched an AI agent marketplace to expand the ecosystem. Mike Sicilia, co-CEO of Oracle and leader of the applications side of the company, said Fusion AI agents are native to the company's stack and available to customers at no additional costs.

"We have 2,400 customers already leveraging Oracle AI embedded inside our applications today," said Sicilia. "They are live across a large variety of industries. Fusion applications are quickly becoming a collection of AI agents."

Today's reality: Oracle applications revenue tops infrastructure

Given the growth rates of Oracle's cloud infrastructure business, up 55% in the first quarter, it's easy to forget that the company even has an applications business. After all, Oracle has been all about Nvidia GPUs, now AMD too, and building capacity for customers like OpenAI.

Here's the comical part. In the first quarter where Wall Street went bonkers over remaining performance obligations, Oracle's cloud applications revenue was still larger than infrastructure. Oracle's applications business, primarily Fusion Cloud ERP and NetSuite, delivered first quarter revenue of $3.8 billion, up 11% from a year ago. Fusion Cloud ERP was up 17% and NetSuite up $1 billion. Oracle's SaaS revenue for the quarter was $3.8 billion. Cloud infrastructure revenue in the first quarter was $3.35 billion.

Going forward, Oracle's revenue mix will change toward infrastructure (at lower margins for a bit). But for now, don't discount applications. For fiscal 2025, Oracle cloud applications revenue was $14.3 billion compared to cloud infrastructure revenue of $10.23 billion.

Another point for your Oracle future file. Oracle's applications business will have to more than pull its weight to fund the OCI investment.

Industry and process focused

Sicilia spent a good bit of time talking about Oracle Fusion and its industry plays. Oracle has its core ERP and HCM systems, but company has a penchant for targeting industries and regulated processes.

Healthcare is the obvious play and one of CTO Larry Ellison's passions. Oracle is planning to automate entire healthcare ecosystems.

"We've got dozens of AI agents live across our health ecosystem today, many more planned. We're looking at chart review care navigation, clinical decision support, patient risk predictions, preventative care and many more. In fact, our next-generation AI EHR is now generally available," said Sicilia.

To Sicilia, the more regulated an industry is the more Oracle can play a role. He said in the next year, Oracle will have 125 AI agents live in banking and insurance.

Oracle is aiming to automate multiple industry ecosystems. The company is projecting fiscal 2030 revenue of $225 billion with a compound annual growth rate of more than 31% over the next five years. For earnings, Oracle is targeting $21 cents a share in annual earnings by fiscal 2030.

An Apple-ish approach to the stack

In many ways, Oracle has lifted the Apple approach with integrated hardware and software. With tight integration, Apple has argued it can provide a better experience.

Oracle's integrated stack between infrastructure, database and applications is a similar story. AI is juicing that integrated stack approach.

Clay Magouyrk, co-CEO of Oracle overseeing OCI, laid out the integrated stack strategy.

"We do applied technology. We do applications. We use our tech, and we also build the tech. We build data centers. We train models. We build databases. We build code generators. We do all this stuff to enable the creation of applications and then we actually create the applications. The database is now inextricably linked to our AI strategy and inextricably linked to our application strategy. So all the pieces at Oracle now are fitting together."

• Holger Mueller: Oracle Makes It 4 for 4: Oracle Database@AWS Is Generally Available | Oracle Solves Decades-Old DBMS Challenges and Future-Proofs Oracle Database
• Stargate ramps as OpenAI, Oracle, Softbank outline 5 new data centers
• Oracle names Magouyrk and Sicilia co-CEOs as Catz moves to Vice Chair

Magouyrk said it wasn't that long ago that the integrated approach looked like a disadvantage because the company was doing too much. The applications and infrastructure approach gives Oracle a built in feedback wheel to continuously improve.

Mark Hura, Oracle President of Field Operations, said the company is engaging companies with its entire portfolio. The argument is that Oracle can take away the enterprise integration pain.

"For our customers to be able to consume and bring outcomes that deliver real value, where they're not focused on integrating solutions, they're not focused on deploying differentiated capabilities, they're not worrying about where their data is moving and how is it secured or not," said Hura.

Miranda said Oracle is threading the needle where customers are able to spend less on IT, but spend more with Oracle. "The customers are actually spending less. They're spending more with Oracle because of the single stack advantage and all the engineering putting all the pieces together instead of customers having to stitch it together," said Miranda.

Sicilia added that the Oracle ambition is have the world run on its integrated stack. "What does it look like beyond 2030? I think we're rapidly heading towards self-implementing self-learning, self-healing systems," he said. "What does it mean in terms of Oracle? It's hard to say. I can't imagine that there's anything negative to say about it because I like our chances of being able to deliver that full system ecosystem."

Constellation Research's take

Constellation Research analyst Michael Ni made the following points about Oracle's strategy.

• Vertically AI integrated. "Oracle moved from 'catch-up' to a credible top-tier option for decision-centric, governed agents. Especially for enterprises already standardized on Oracle data and apps. AWS and Microsoft remain broader ecosystems, and Google is strong on tooling and AI platform and partners. Oracle’s differentiator is vertical integration from database to governance that expands to private data to apps with agent lifecycle controls promising to simplify management and decrease costs."
• The stack not the tools. "Oracle’s embedding move is classic Larry Ellison. Centralize complexity by delivering an “engineered stack” that brings tight integration, delivered semantics and data models, and governed context. This helps Oracle own the applications context and actions differentiates Oracle’s offering from other hyperscalers. While AWS and Google sell you the toolkit, Oracle ships the machinery already wired for output. That’s provides performance and extensibility, but at a tradeoff to portability."
• Differentiated. "Oracle’s strategy is differentiation through vertical integration, not head-to-head competition on horizontal cloud scale. While Oracle Cloud Infrastructure (OCI) is already positioned has highly competitive from a price-performance perspective, Oracle is doubling down on the enterprise data–to–decision stack by connecting applications, context and ability to act, where governance, trust, and application context matter more than raw hyperscale capacity."
• AI peanut butter and jelly. "Oracle's dual investment in AI and infrastructure are not two bets. Those investments are one play: control the cost curve and own the decision loop as a data and AI foundation for enterprise AI. Ellison eyes the large hyperscaler market and knows Oracle can differentiate if it can manage the margin hungry AI solutions that is increasingly creating cost as fast as it creates value for the application level. You can’t win the AI platform war without owning both the brains (agents) and the body (infrastructure)."

Welcome to the context chorus: There’s no AI without context

Larry Dignan

Editor in Chief of Constellation Insights
Constellation Research

Larry Dignan is Editor in Chief of Constellation Insights at Constellation Research, where he leads editorial coverage focused on enterprise technology, digital transformation, and emerging trends shaping the future of business. He oversees research-driven news, analysis, interviews, and event coverage designed to help technology buyers and vendors navigate complex markets with clarity and context. ...

Read more

Don't look now, but enterprise technology vendors have a new fascination with context engineering to give AI agents the data and insights to go from pilot to production. You can almost see "context" trending as a theme the more enterprise vendors talk.

Elastic on its investor day earlier this month talked extensively about context. Elastic has enterprise search, observability and security platforms that are leveraged to ingest data from multiple sources, process it and create AI agent experiences and workflows.

Ashutosh Kulkarni, CEO of Elastic, led the charge with a focus on context. Kulkarni said that large language models (LLMs) have transformed what you can do with unstructured data and represent the new enterprise operating system.

• Agentic Automation Doesn’t Fail at Reasoning. It Fails at Inputs.
• Elastic's Agent Builder is generally available, Elastic Workflows in tech preview

To make that AI operating system work though, you need context and data. "To use LLMs within the enterprise, you have to provide them with context. You have to provide them with relevant data to be able to address the problem that they're trying to address. So, AI fundamentally depends on data, being able to have access to it. And relevance is key to making any AI system worthy, production grade," said Kulkarni.

In other words, data matters, but context and relevance may matter more.

Kulkarni made the case that Elastic is set up as a premier platform for context engineering. Context engineering is a set of technology approaches designed to give LLMs the right data and right tools at the right time to do job accurately.

Elastic launched its Agent Builder and the Elastic Inference Service, which will give users access to the company's embedding, retrieval and reranking models. Kulkarni said context and relevance needs more than just a vector database. Elastic also acquired Jina AI to build out its multimodal models for looking at documents.

Ken Exner, Chief Product Officer at Elastic, said the company's long-running focus on relevance has given it the ability to pull context out of data and present it to other systems.

Exner said LLMs have to be grounded with the right data and context to be accurate. Agentic AI only ups the ante. "Now that an AI agent is performing an action, it's doing a task and potentially doing it badly," said Exner. "The consequence of not grounding that LLM or that agent, not giving it the right context could be disruptive and damaging. I think relevance matters so much in the age of AI. Context engineering as a concept is going to be talked about constantly as people move forward with agentic AI because it's all about having relevance. It's all about having context. That's what you need to do AI correctly."

Elastic's discussion of the importance of context was a notable theme by itself, but then the context chorus revved up.

At Dreamforce 2025, Salesforce CEO Marc Benioff's keynote had a hefty dose of context. The pitch for Agentforce is that the Salesforce platform has your customer data and can personalize at scale. "AI alone is not enough. It's not enough just have an LLM. It needs (Agentforce) capabilities to connect it, to give it the context, to give it the guardrails, and all of the critical pieces," said Benioff.

The argument for Agentforce, according to Benioff and a bevy of Salesforce executives, is that you need the customer data, the ability to navigate unstructured and structured information, and trust to bring context to every conversation.

Dreamforce 2025

• Salesforce sets $60B revenue target: Will unlimited Agentforce plans drive growth?
• Salesforce CEO Benioff: Context is king
• Salesforce makes its Agentforce 360 case to be your AI agent platform
• Salesforce's acquisition of Apromore highlights how process intelligence, agentic AI converging
• Salesforce expands OpenAI, Anthropic partnerships, eyes Agentforce everywhere

The context engineering topic is one worth pondering as enterprises start designing different types of agents with unique architectures. Salesforce updated its enterprise agentic architecture whitepaper as did Google Cloud. Both are worth a read since each AI agent type will have a unique flow and design. One obvious takeaway is that context will be critical throughout that AI agent design.

Workday Rising also featured a bit of the context thread, largely because Workday is building its Data Cloud to combine the data on people and money. “We're building the best AI agents for HR and finance that deliver real business value. Our data has context, and we're deeply embedding AI into HR and finance processes,” said Workday CTO Peter Bailis. “We're shifting from the system of record for people and money to a system of action for people and money that understands their customer businesses, understands what they need and helps them reimagine what work gets done.

The working theory is that most vendors are going to talk about context just as much as they do data, AI and AI agents. For instance, Constellation Research analyst Michael Ni recently attended Teradata's conference and reported the following:

Joining this month's parade of vendors focused on context, Teradata believes there's no AI without context. For them, context isn't just data & it's the metadata, business logic, and domain know-how that make AI decisions relevant and reliable. Over 95% of enterprise AI projects fail, they argue, because models are built in isolation from business context. Without that grounding, even the best algorithms can’t deliver the accuracy or explainability needed in real-world, regulated environments.

To close that gap, Teradata is turning decades of decision analysis experience into domain and industry knowledge models. These include pre-built layers of business logic, KPIs, and rules from managing customer lifetime value to financial services metrics that give AI agents real context from day one. Their context intelligence framework captures how industries actually operate, so organizations don't have to start from scratch. The goal: help teams build agents faster, with enterprise-grade performance, governance, and trust already built in.

Big bank CEOs, CFOs riff on state of AI exuberance

Larry Dignan

Editor in Chief of Constellation Insights
Constellation Research

Larry Dignan is Editor in Chief of Constellation Insights at Constellation Research, where he leads editorial coverage focused on enterprise technology, digital transformation, and emerging trends shaping the future of business. He oversees research-driven news, analysis, interviews, and event coverage designed to help technology buyers and vendors navigate complex markets with clarity and context. ...

Read more

The big banks and financial giants have reported earnings and across the board they were strong due to increased M&A and trading volume. That said it’s worth noting some key comments about the AI infrastructure buildout. Here’s a look at what financial sector CEOs and CFOs are saying about the AI market.

Goldman Sachs CEO David Solomon said on the company’s fiscal third quarter call:

“There is no question that there is a fair amount of investor exuberance at the moment with U.S. equity markets consistently hitting record highs over the last several months. Much of this has been fueled by a tremendous amount of investment in AI infrastructure, which has driven significant capital formation. But as students of history, we know that following periods of broad-based excitement around new technologies, there will ultimately be a divergence where some ventures thrive and others falter. While I feel good about the forward outlook on balance, the market operates in cycles and disciplined risk management is imperative. We are especially vigilant in times like these to proactively manage risks.”

• Goldman Sachs eyes next AI transformation from position of strength
• Watercooler debate: Are we in an AI bubble?
• Enterprise AI: It's all about the proprietary data

JPMorgan Chase CFO Jeremy Barnum talked about his opinion on the AI boom and current market. He said: “The risk is because of how incredibly overwhelming the AI theme is for the whole marketplace right now and all the various effects that it's having in terms of equity market performance, MAG 7, data center build-out, electricity costs, like it's an overwhelming thing. And I think for us, running a company of this type, we need to make sure we stay anchored in like facts and reality and tangible outcomes. We’re putting a lot of energy into AI. A lot of people are spending a lot of time on it. We're spending a lot of money on it. We have very deep experts.”

Barnum continued:

“We've been doing it for a long time, well before the current generative AI boom. But in the end, the proof is going to be in the pudding in terms of actually slowing the growth of expenses. And so, what we're doing is kind of rather than saying you must prove that you're generating this much savings from AI, which turns out to be a very hard thing to do. It’s hard to prove and might at the margin result in people scrambling around to use AI in ways that are actually not efficient and that distract you from doing underlying process reengineering that you need to do. What we're saying instead is let's just do old-fashioned expense discipline.”

Citigroup CEO Jane Frasier had a similar theme on the bank’s third quarter earnings call.  She said: “The macro environment reflects the global economy that's proved more resilient than many anticipated. The U.S. continues to be a pace setter, driven by consistent consumer spending as well as tech investments in AI and data centers. That said, there were pockets of valuation frothiness in the market. So, I hope discipline remains.”

• Citigroup AI agent pilot underway for 5,000 workers

The AI buildout is also a theme abroad, notably China. Frasier said:

“In Asia, China's domestic spending has slowed. However, the investments they are making in technology are staggering, and the world should take notice. India's fundamentals of the young tech-savvy labor force and robust domestic consumption continued to drive high growth there. But in Europe, structural challenges still need to be dealt with for the continent to escape this low growth cycle.”