AI agents move from demos to deployment at enterprise scale – News

Something’s shifting under the surface of enterprise tech. For years, companies have tried to streamline operations, connect siloed systems and bring consistency to customer engagement, but cracks keep showing. Fragmentation still wins. The future will belong to those who rethink architecture from the ground up, especially as AI agents begin reshaping how work gets done.

That rethink is already taking shape, according to George Gilbert, principal analyst for data and AI at theCUBE Research. He sees the rise of intelligent agents as the next epoch in enterprise software — one that could surpass even the relational database in its impact. In his view, Salesforce Inc. stands apart as the first major vendor with all the components needed to unify customer data and dissolve long-standing silos.

“In the agent revolution, it’s really the biggest transformation in enterprise software that we’ve gone through since the relational database, and maybe bigger than that,” Gilbert said. “Salesforce is the first major application vendor or platform vendor of any stripe who has all the pieces put together to be able to start unifying … all of the customer-related data and processes. It’s like no more islands.”

Gilbert, alongside fellow analysts John Furrier and Gemma Allen, spoke with chief information officers, presidents and digital transformation executives during Dreamforce. The conversations revealed how AI agents are moving from proof-of-concept demos to production-grade deployment, raising new challenges around data fluidity, governance and measurable business value. (* Disclosure below.)

Here’s the complete video interview with George Gilbert, John Furrier and Gemma Allen:

Check out three insights you may have missed from theCUBE’s coverage of the event:

Insight #1: Data fluidity, observability and governance turn agent demos into enterprise reality.

The foundation of any agentic system — and of reliable AI agents — is unified, accessible data, according to Rahul Auradkar, executive vice president and general manager of unified data services at Salesforce. Built on open standards, Salesforce’s architecture encourages ecosystem-wide participation. It supports a transparent data layer that can handle structured and unstructured formats, which is key to real-time semantic understanding.

Salesforce’s Rahul Auradkar talks with theCUBE about the company’s vision for agentic AI.

“We believe in this idea known as data fluidity or fluidity … the ability for the context to be available for the agent in the moment,” he said during the event. “We have built Data 360 from the ground up on standards … we are an open and extensible platform. That sort of an ecosystem vibrancy is needed for data fluidity that allows us to provide intelligent context in the moment.”

Agentic systems don’t just automate; they also orchestrate the behavior of AI agents. This functionality requires a new control plane for decision intelligence, according to Gilbert. As Salesforce’s infrastructure matures, it signals a shift in how enterprises approach governance, performance management and visibility at scale.

“There is a testing center, which is like an observability platform,” Gilbert noted. “It captures the reasoning traces and the data and the decisions and the actions … that each agent takes. It captures that fine thread. It’s the new big data.”

Enterprise-grade control systems are now essential as companies move from demos to large-scale agent deployment. While artificial intelligence adoption is driving record IT growth, too many organizations focus solely on cost-cutting, undermining long-term value. Salesforce’s customer relationship management capabilities provide customers with a path to unlock revenue growth and enhance efficiency, according to Crawford Del Prete, president of International Data Corp.

“The magic happens when you can unlock revenue growth while you’re reducing costs,” he told theCUBE. “Today, what I’m seeing — what’s troubling — is too many companies are only focused on the expense side of the equation with AI, and that’s where I think Salesforce has an asset because they, through their CRM, can figure out the use case around making your leads better, warming up [and] training your Salesforce … in a new way. They can unlock that message and use AI to drive growth.”

Here’s the complete video interview with Rahul Auradkar:

Insight #2: The agentic enterprise moves from tech vision to growth strategy.

The shift to an agentic enterprise is central to Salesforce’s long-term growth strategy, according to Robin Washington (pictured), president and chief operating and financial officer of Salesforce. The company’s 50 by FY30 plan blends margin discipline with service-driven revenue, and leaders see agentic transformation as the most critical driver of financial value in the years ahead.

PepsiCo’s Dave Dohnalik and Johannes Evenblij Garza talk with theCUBE about how Agentforce has helped shape the company’s customer relationships.

“We really see a once-in-a-lifetime opportunity with the agentic enterprise to really accelerate our customers’ adoption of AI, which will fuel our growth and improve our customers’ success,” Washington told theCUBE. “We want to help our customers leveraging technology solve their greatest problems.”

That vision is mirrored by clients such as PepsiCo Inc., which is deploying Agentforce to strengthen customer relationships and unify data across sources, according to Dave Dohnalik, senior VP of technology strategy and enterprise products at PepsiCo, and Johannes Evenblij Garza, SVP of global sales and marketing transformation at PepsiCo. The goal is to move from a reactive model to proactive, real-time engagement across 6 million customers, using AI agents to handle repetitive tasks and speed up sales workflows for frontline reps.

“Our own frontline people … typically they go to 15, 20 stores a day,” Dohnalik said during the event. “They really don’t have time to understand, ‘Hey, what do you need, customer?’ Now they go and they can see in a mobile app-digested brief on exactly what is needed, and then they talk and they can discuss it in just a couple of minutes with the customer. It works really well.”

Here’s the complete video interview with Robin Washington:

Insight #3: AI agents amplify the human touch across cities, commerce and classrooms.

The City of Kyle, Texas, is using Agentforce to speed up citizen services and deliver measurable results, according to Jesse Elizondo, assistant city manager at the City of Kyle. By automating low-touch requests with AI agents, the city cut average resolution times nearly in half and saved more than the entire cost of its 311 implementation, freeing staff to focus on high-value work for its 70,000-plus residents.

City of Kyle’s Jesse Elizondo talks with theCUBE about modernizing city engagement while saving taxpayer money.

“One of the things that we did when we first utilized AI was that we were getting thousands and thousands of calls during the day,” he told theCUBE. “But one of the use cases we’ve used was during the second and third shifts. After 5:00, we don’t get a tremendous amount of calls. It wasn’t enough to validate a full-time employee to take those calls. We have this system, this AI system that sits on top of our CRM, our 311. It knows everything about service requests. It can do all kinds of intake.”

SharkNinja Operating LLC is also embracing agentic AI to deepen consumer relationships and create more intuitive experiences, according to Velia Carboni, chief information officer of SharkNinja. By consolidating its two brands into a single platform, the company is shifting from transactional selling to long-term customer connection — driven by a belief that intelligent agents represent the next major leap in business evolution.

“I think we can’t be afraid about technology changes,” Carboni said during the event. “This is the next evolution of, ‘How do we create better consumer experiences, make them more streamlined and really take those experiences to the next level? ‘ I encourage everyone to jump in and not fear it, but embrace it.”

DeVry University is bringing the agentic model to education by integrating AI-driven personalization across the entire student journey, according to Chris Campbell, chief information officer of DeVry. With many students needing support outside traditional hours, DeVry uses agents to deliver real-time assistance and to track measurable value for both learners and the institution.

“We are already seeing benefits that we can tie directly [return on investment] to both persistence and retention of students, both key things for delivering value in their lives and, frankly, value to the university,” Campbell explained during the event. “We have leveraged agents on our new DeVryPro platform, which is a continuing professional education platform that delivers stackable credentials and is AI-powered.”

Here’s the complete video interview with Velia Carboni:

Want more insights from Dreamforce? Check out these interview segments:

• Jayesh Govindarajan, EVP of software engineering at Salesforce, and Christophe Coenraets, SVP of technical audience relations at Salesforce, talk with theCUBE about how Agentforce is redefining AI-driven software creation, enterprise vibe coding and the craftsmanship behind modern development.
• Holger Mueller, VP and principal analyst at Constellation Research Inc., and  Raphaelle d’Ornano, founder and chief executive officer of Decoding Discontinuity, talk about Salesforce’s new focus on AI and agentic orchestration.
• Nancy Xu, VP of AI and Agentforce at Salesforce, talks with theCUBE about the agentic lifecycle and how Salesforce is shifting who gets to shape AI inside the enterprise.

To watch more of theCUBE’s coverage of Dreamforce, here’s our complete video playlist:

(* Disclosure: TheCUBE is a paid media partner for Dreamforce. Neither Salesforce, the sponsor of theCUBE’s event coverage, nor other sponsors have editorial control over content on theCUBE or News.)

Photo: News