Shaping the next era of agentic AI at Think 2026

4 p.m. EDT | May 6, 2025

IBM executives used the company’s Think conference this week to argue that quantum computing is moving from experimental science to commercially relevant applications.

The announcements came on the last day of the conference, as technology companies attempt to prove that quantum systems can solve commercially meaningful problems after years of scientific breakthroughs and billions in investment. On stage with its partners, IBM unveiled new results that executives said pushed quantum systems beyond research demonstrations and into practical work involving drug discovery, materials science and fusion energy.

“Useful quantum computing is here right now,” Director of IBM Research and IBM Fellow Jay Gambetta said on stage.

He framed industry’s next phase around what IBM calls “quantum-centric supercomputing,” an architecture that combines CPUs, GPUs and quantum processing units (QPUs) into tightly integrated systems. Instead of replacing classical computers, the systems divide calculations between conventional and quantum hardware, depending on which machine handles a specific task best.

IBM pointed to a series of early scientific use cases to support its argument that quantum computers are beginning to tackle problems that strain conventional systems. One of the keynote demonstrations came from researchers at Cleveland Clinic, who described using IBM quantum systems to model increasingly large biological molecules tied to pharmaceutical research.

Kenneth Merz, who leads a lab at Cleveland Clinic that collaborates with IBM on hybrid quantum-classical computing models for biomedical research, said that his team began with relatively small molecular calculations before scaling up to protein systems with more than 12,000 atoms.

“We did this, and we thought we were really going big, but we quickly realized we could have been two or three times larger,” Merz said.

Researchers used a technique called SQD, short for sample-based quantum diagonalization, to estimate molecular behavior at scales that classical methods increasingly struggle to handle efficiently.

The company also highlighted a recent project involving a newly synthesized “half Möbius” molecule, a structure shaped like a twisted loop that does not naturally occur in nature. IBM researchers used quantum systems to simulate the molecule’s electronic structure and compare it with experimental imaging data. Gambetta said the results showed quantum hardware helping interpret real laboratory measurements rather than isolated benchmark exercises.

IBM and Oak Ridge National Laboratory researchers also talked about a project to use quantum systems to study fusion reactor fuels, a key challenge in efforts to develop commercially viable nuclear fusion energy. The effort combined AI, high performance computing and quantum processors to model the chemistry involved in tritium production. Scientists said the chemistry remains difficult to model accurately using conventional computing methods alone.

Sarp Oral, Section Head for the Advanced Technologies Section and a Distinguished Research Scientist at the National Center of Computational Sciences (NCCS) Division of Oak Ridge National Laboratory, told the session that the project combined AI, high-performance computing and quantum processors into a coordinated workflow aimed at accelerating scientific research.

“What we are trying to build in this workflow is to bring together different computing paradigms,” he said.

One of the session’s biggest announcements came from quantum computing startup Q-CTRL, which shared that it had achieved what it called “practical quantum advantage” using IBM quantum hardware. The company’s CEO and Founder, Michael J. Biercuk, said that the companies solved a commercially relevant materials science problem using a 120-qubit IBM quantum system: the quantum workflow completed the calculation in roughly two minutes, while a comparable classical simulation took more than 100 hours on a computing cluster.

“That’s a 3,000 times speedup owing to access to quantum computers,” Biercuk said.

12:30 p.m. EDT | May 6, 2025

IBM’s “Accelerate AI ROI with Hybrid Cloud” keynote at Think 2026 centered on a question that has grown harder to avoid: after more than a year of AI experimentation, why is so little of it in production?

IBM SVP of Infrastructure Ric Lewis didn’t sugarcoat the situation. The industry, he said, is fielding a lot of strategic questions at once—which foundation model would win, whether companies would build or rent AI capabilities—while the “SaaSpocalypse” raised longer-term questions about vendor survival. “With all those questions and the unprecedented speed of change in the industry, you might be a little bit stressed out,” Lewis told the audience. His answer was to cut through it. “To best position yourself, you can build an underlying infrastructure that’s flexible and ready for this dramatic rate of change,” he said. Data, he argued, is either a gold mine or a landfill—and architecture is what determines which.

“The world is hybrid,” he said. “Now you’ve got to deal with that.”

Skyla Loomis, GM of IBM Z Software, pushed back on the idea that mainframe AI is something to hand off to the infrastructure team and forget about. “Activating AI on the mainframe is a strategic business decision,” Loomis said. “It is about removing friction, unlocking greater value from your applications and data, while protecting the core of trust and cost efficiency.”

IBM had a roster of announcements to back that up. IBM Bob Premium Package for Z—the evolution of watsonx Code Assistant for Z, now in tech preview—is a mainframe development tool, with early users reporting 10x productivity gains. IBM Z Database Assistant brings that AI productivity to additional personas on the platform. zSecure Secret Manager handles certificate lifecycle management on IBM Z, connecting to IBM Vault for visibility across the environment.

Loomis was joined on stage by Jikin Shah, Global CTO of Royal Bank of Canada. RBC’s hybrid cloud now supports all business-critical workloads, with more than 45% of workload running through it. That foundation has enabled AI products including a proprietary trading model called ATOM (Asynchronous Temporal Model), trained on billions of internal client transactions and RBC Clear, a cloud-native cash management platform.

Getting there required more than a technology roadmap. Shah described a parallel organizational transformation, including a new head of AI reporting directly to the CEO and a generative AI overlay called RBC Assist that’s now used by more than 50,000 employees. “It requires intentional effort,” Shah said of building RBC’s AI foundation. “It just doesn’t happen by chance.”

As for the tension between speed and compliance, Shah was direct: “Whoever is working in a regulated industry knows there is no magic wand.” Governance, he said, is what gets innovation to production, not what slows it down. “If you don’t have that, you may not be able to take your innovative idea all the way up to production and make it available for your clients and stakeholders.”

Sam Werner, GM of IBM Storage, made the case that for most organizations, data is the bottleneck AI keeps running into. “Most organizations don’t really have a data foundation that’s ready yet to deploy AI at scale,” he said. “It’s this data foundation that ultimately makes the difference for an organization, whether they get their AI into production, or [whether] they get stuck in this endless mode of AI pilots.”

Werner named two failure modes that keep coming up in client conversations: data that is too fragmented and siloed to feed AI at enterprise speed, and governance that is bolted on too late to matter. “How are you going to deploy AI at scale, or bring agents into your model and allow them to have access to the data, if you don’t have the governance required to ensure the data is safe and protected?” he said.

IBM’s pitch is a full-stack fix: IBM watsonx.data for unified data, with governance built in from the start, and IBM Fusion for availability and performance—with intelligent data caching that brings data to the AI as needed, without requiring organizations to re-platform.

Werner was joined by Riccardo Perrotta, Head of IT Service, Operations and Infrastructure at Unipol Group, one of Italy’s largest insurance conglomerates. His team runs IT for 32 companies across insurance, healthcare, payments and financial services out of a single centralized hub. “A standard operating model is not good for everyone,” Perrotta told the audience. “Each company has different needs. Each company has a different regulatory restraint.”

Unipol’s answer is NAMI (short for Next Automation Monitoring Insurance), an AI-powered IT operations platform implemented in June 2025. It monitors events across technical and business systems in real time and routes structured incident reports rather than alarms. Rather than a single large model, it runs specialized models matched to specific tasks. “We do not want one model for everything,” Perrotta said. “We want the right model for the right job.” Running on IBM Fusion and expanding to IBM z17, NAMI has outgrown its original scope: “NAMI is not only an AI solution,” Perrotta said. “NAMI is a platform to manage the AI.”

Hillery Hunter, IBM’s CTO of Cloud and Datacenter Products, closed on the question of control: who actually owns the infrastructure your AI runs on, and what happens when that changes? “There’s one persona that stands out to me,” she said. “That’s the enterprise architect. And I think 2026 is their year.”

Sovereignty, she said, means being able to control your operations, know that you have a solid grip on your data and know what is going on with the AI in that environment.

IBM has launched two products to make that real: IBM Sovereign Core, for deploying and managing sovereign AI environments, and IBM Cloud Sovereignty Risk Profile, for audit-ready evidence of compliance.

“AI plays a role, therefore, in entering this era with confidence that you are making infrastructure decisions that you can manage and deliver the business outcomes that this era of speed requires,” Hunter said. “Infrastructure has never mattered more.”

10 a.m. EDT | May 6, 2025

For companies to succeed in executing AI at scale, it comes down to the people who use it, not the technology itself. That was the core theme in “How enterprises excel in the AI era,” a Think 2026 keynote session featuring real-world examples of AI transformation from New York Life Group Insurance, Warby Parker, The Walt Disney Company and IBM.

“When we think about AI transformation, we think, ‘Oh, that’s a technology project,’” said Scott Berlin, SVP and Head of New York Life Group Insurance. “But the reality is transformation is a people project.”

He described the technology transformation at New York Life as being grounded in its workforce, from beginning to end. “We’re redefining the interaction model between our people, the technology platforms, the agent capabilities. It’s an end-to-end transformation of the business.” With a team of more than 3,500 employees, Berlin said he is looking to change the job description of every role, deploying AI across the entire business workflow so that workers are able to spend more time where it counts: “providing empathy [to customers] in those moments that matter.”

In the same vein, IBM has driven AI growth at the company by starting with the people first, focusing on how teams adopt and adapt to the tools to see what works and what doesn’t.

Jonathan Adashek, SVP of Marketing and Communications at IBM, cited a McKinsey study on how organizations are using AI. Nearly 90% of the companies surveyed said they have deployed AI in at least one business function, but 94% of those respondents claimed they are not seeing significant value in those deployments. The way Adashek sees it, transformation begins with a behavioral change in the people. “[This] gap between expectations and outcomes is the largest leadership challenge of the decade,” Adashek said. “Technology implementation without change management won’t work.”

To bridge the gap, Adashek shared his own experience in leading that change management within IBM’s Marketing, Communications and Corporate Social Responsibility (MCC) organization. In prioritizing outcomes, and the people behind those outcomes through continuous feedback loops, new workflows and reskilling, MCC was able to unlock millions in cost savings, on top of seeing faster time to market, stronger buying signals, thousands of hours saved and tighter alignment between marketing and sales. “This is not just about putting AI on top of old ways of working,” Adashek said. “It’s about putting AI at the center of the organization and building around that.”

For Susan Doniz, Chief Information and Data Officer at The Walt Disney Company, AI transformation has a clear north star: “The adoption and success is all about the humanity.”

She explained how, in the past, Disney’s business model operated much differently than it does today, with a third party between the company and its customers. To see a movie, she said, fans had to go to the cinema; to see something on television, people had to go through a broadcaster. “There was always somebody in between us and our fans,” Doniz said. Now, operating as a direct-to-consumer business, Disney is able to deliver “more experience, more directly, more personally,” tailoring the experiences to customers’ evolving needs. “Our first job is always to bring joy and happiness to our fans, and now, without this middle person, it means we have a direct line of sight to bring this joy to everybody.”

But to get there, “You can’t do things to people—you have to do it with them,” she said. “You have to put the power in their hands.”

Doniz acknowledged that people are feeling overwhelmed “as a consumer, as a viewer, as an employee” in this era of fast change. To alleviate that pressure, Disney hosts “art of the possible” sessions for its employees—workshops where teams come together to simplify their work, solve real problems through design thinking, and set “big, hairy, audacious goals.”

For Doniz, there’s no holy grail to getting it right, and there are “hard yards of continuous improvement” ahead. But the mission remains central: “What’s written on the piece of paper is not as important to what [we are] trying to accomplish,” Doniz said. “If we do this together, as the Lion King said, we will go very far together.”

8 a.m. EDT | May 6, 2025

Companies are pushing artificial intelligence agents beyond chatbot demos and into the daily grind of recruiting, healthcare staffing and customer service, executives shared on day two of Think 2026.

During a keynote titled “Architecting the AI-first enterprise” executives from IBM, Amazon Web Services, Providence Health, Away and Pearson described how companies are embedding AI systems into operations as pressure mounts in a competitive business environment to move faster and cut costs.

“By 2030, 50% of operational decision making will be done by AI,” said Neil Dhar, Senior Vice President at IBM Consulting. Dhar added that companies would still need “a human in the loop” to apply judgment and oversight.

Executives said businesses now face shrinking product cycles, rising customer acquisition costs and competition from startups built around AI from day one.

Consumer brands are trying to keep up with customers who expect faster, more personalized experiences, executives at the conference said. “The customer doesn’t shop in channels anymore,” said Jessica Schinazi, CEO of luggage company Away. “They think about whether the brand is showing up for them.”

Away started as a direct-to-consumer luggage startup. Now the company sells through partners including Amazon and Nordstrom while expanding into what Schinazi described as “agentic channels,” or AI-driven recommendations and shopping systems.

She said AI tools now help the company analyze customer reviews, synthesize research and test ideas at far greater speed. Work that once took weeks can now happen overnight. “We are able now to spot emerging needs earlier,” Schinazi said. “We are able to pressure-test the concepts much faster than the competition.”

Healthcare systems are moving in the same direction. Leaders at Providence Health & Services said the organization built an AI-powered recruiting and transfer system named Rita to automate administrative work and accelerate hiring across its hospital network.

“We needed to move faster, respond to candidates more quickly, maintain compliance and free our recruiting teams to focus on what matters most,” said Carol McDaniel, Vice President of Talent Acquisition at Providence Health.

McDaniel said Providence partnered with IBM to develop the system, which reduced leadership processing time by 90% and helped fill some positions 12 days faster. The goal, she said, was to move healthcare workers into patient-facing roles more quickly.

These systems come with their own challenges. “Security’s got to be there from the start,” said Greg Pearson, Vice President of Global Sales at Amazon Web Services. “It’s not something you add after the fact.”

Pearson said IBM Consulting’s AI agents run on Amazon Bedrock, AWS’s managed AI platform. He also pointed to work with the insurance company Fortitude Re, done in partnership with IBM, where AI systems reduced claims processing time from more than six weeks to roughly 10 days.

As companies deploy larger numbers of AI agents into sensitive operations, executives also described efforts to verify whether those systems can reliably complete assigned tasks. At Pearson, the education and testing company, executives are working with IBM on systems designed to test and credential AI agents before deployment into business workflows.

“We believe that agent skills cannot be assessed the same way we assess humans,” said Dave Treat, CTO at Pearson. “Trust depends on how they perform inside a defined business workflow.”

Treat said the process evaluates whether AI agents can follow instructions, apply technical knowledge and complete tasks under real-world conditions before participating in operational work.

Even as executives described broader adoption of AI systems, several speakers argued that human judgment will become more important as automation spreads. “The skills that will rise in value are the ones that machines can’t fully own,” Treat said. “Judgment, empathy, accountability, communication, leadership. Humans will still define the standard.”

4:30 p.m. EDT | May 5, 2025

“There’s an emerging divide in AI, and it’s accelerating day by day,” said Rob Thomas, IBM SVP of Software and Chief Commercial Officer, addressing a packed room Tuesday at Think 2026. Sharing the stage with several of his IBM colleagues and select IBM clients, he emphasized the importance of the AI ecosystem—the practice of adopting AI with an operations mindset—which he calls the “AI Operating Model”—and stressed that the companies that win are those that implement a holistic model. “The question isn’t, ‘Which AI should you choose?’” he said. “It’s, ‘How do you operate AI across everything you already have—and everything that’s coming next?’”

Thomas compared AI with the advent of another revolutionary technology: electricity. “Electricity led to the light bulb, and that improved factories,” he said. But it was another use case that proved far more disruptive 30 years later: the assembly line. “The assembly line changed productivity forever; that’s when industry changed,” he said, noting that enterprises should think of AI not in terms of light bulbs, but in terms of assembly lines.

When it comes to building an AI Operating Model that integrates multiple systems and processes, he said, what holds many companies back is fear of failure. But while it can be daunting to reconceive the whole assembly line, as it were, he said it’s supposed to be hard. “It comes back to the first principle of enterprise technology: it’s never going to be as simple as one platform or one cloud or one model,” he said. “The opportunity is not whether you adopt AI, but [rather] which side of the divide you’re on.”

Jay Kreps—Cofounder and CEO of the major data streaming platform Confluent, which was recently acquired by IBM—explained how AI Operating Models made all the difference. During his time as an engineer at LinkedIn, where he was technical lead for their relevance and data systems, part of his team’s responsibility involved running batch processes. “By the time you get your data, the customer’s gone,” he said. “We had to really be able to bring together the whole ecosystem around real-time data.” He tackled the issue in part by making the company’s intelligence open source; later, at Confluent, he would add hybrid cloud to the mix.

The results of Kreps’ AI Operating Model speak for themselves. Marriott, a Confluent client, leveraged Confluent technology to “bring together all the data they had about customers—all the personalization, loyalty programs, marketing—really unite everything that they knew about the people who were staying with them,” Kreps said. “[We helped them] optimize that experience.” To understand the efficacy of this initiative, one need only learn that this overhaul of Marriott’s customer personalization programs added USD 250 million in incremental revenue, according to Kreps.

12:30 p.m. EDT | May 5, 2025

When world tennis phenomenon Andre Agassi was set to play Andy Roddick for the first time in Miami, he did what any Grand Slam champion would do before the match: he stopped his car and stared at a billboard. “I looked up and saw [Roddick] hitting a backhand,” said Agassi alongside Andy Baldwin, Senior Vice President for Offerings and Growth at IBM, at Think 2026. In that one image, he unearthed enough data to formulate a winning strategy against the tennis champion: the strength of Roddick’s grip, the early engagement of his left arm and the power being generated from his legs all contributed to Agassi’s plan of attack.

Whether gathered from a billboard or from real-time training insights, data is valuable. Which means protecting it with proper governance guidelines is paramount.

The AI agent boom has given enterprises a second workforce, but it has also raised new complexities when it comes to managing them efficiently and productively. “By the end of this year, enterprises expect to deploy more than 1,600 AI agents on average—a digital workforce making thousands of decisions every day,” said Baldwin.

So, what’s keeping enterprises from taking advantage of the efficiencies that come with using AI at scale? Baldwin pointed to governance—and specifically, the ability to orchestrate agents and turn their actions into results. Notably, as more employees experiment with AI agents, especially without any guardrails baked in, the potential dangers increase. “7 in 10 executives believe the AI governance they have in place is not fit for purpose, increasing enterprise risk,” Baldwin said.

To demonstrate the power and benefits of directing AI agents, IBM’s Neel Sundaresan, GM of Automation and AI, and Sanah Pallithotungal, Product Manager for Automation and AI, showed off IBM Bob and its agent management capabilities live on stage by asking it to handle a product recall issue at a fictitious company. 

Using watsonx Orchestrate, Pallithotungal was able to train Bob to orchestrate agents, apply governance rules, handle recall processes and manage multiple agents. “Bob accelerates how quickly our teams can start building the agent and respond to real business events,” said Pallithotungal, with Bob and a human reviewing the plan created before deploying. Throughout the development of the new AI agent, Bob is steered by previously established governance rules. “It’s part of how the agent operates from the beginning,” said Pallithotungal. 

Agassi Sports Entertainment is partnering with IBM to enhance the effectiveness of tennis coaches globally. The company’s AI-powered training app—voiced by Agassi’s coach, Darren Cahill—can provide analysis of a player’s technique using a smartphone’s camera. “It allows people to not just see the trajectories of balls, swings, comparatives, but to actually ask [Cahill] questions that aredirectly related to their game, and for him to ask better questions or get more specific about the answer,” said Agassi.

“You’re talking about not putting coaches out of business, but giving coaches in schools, in clubs real access and making sure they’re on point,” Agassi added.

Agassi’s startup is able to move quickly with its growing collection of training data thanks to its partnership with IBM and use of wastonx.ai. But Baldwin noted that most enterprises are sitting on something even more valuable: years of operational data embedded in interconnected systems. “You aren’t starting from scratch,” said Baldwin. “You’re dealing with a lot of complex legacy systems and processes, and any changes you make to that [technology workflow] through AI, you have to ensure the business keeps performing at that level and at that pace.”

Jean-Michel Garcia, CTO for BNP Paribas, joined Baldwin to discuss how the global bank integrated AI agents into its operations without sacrificing security or innovation. Garcia said that BNP Paribas, like other banks, had already been using AI models for some time. “But we saw immediately it would touch every party of the company” in 2023, he said, when various AI models like ChatGPT exploded in popularity.

While BNPP has its own in-house AI model, it is striving for a more hybrid model that takes advantage of different AI models and agents while not being locked to a single vendor. “It was a big shift—not only at the technology level, but [in] the way we think about the future of business,” said Garcia.

Garcia’s ultimate vision? AI that removes the unnecessary tasks that might burden employees today, freeing up their time “to imagine, to build, to design, to think,” he said. “We will be more human inside the company with AI.”

10:30 a.m. EDT | May 5, 2025

“We are on Day Zero of the AI revolution.” That was IBM Chairman and CEO Arvind Krishna’s key message Tuesday morning at the official kickoff for Think 2026 in Boston—and he was quick to clarify what he meant. “Day Zero doesn’t mean it’s yet to come,” he said. “Day Zero is simply that it’s here and now, and you have to take advantage of it now.”

AI-first organizations are already seeing 70% greater productivity improvement and 74% faster cycle times than their peers, he said, and IBM has logged more than USD 4.5 billion in productivity gains applying AI and hybrid cloud across its own operations. Most enterprises, however, are still running AI at the margins, though the window to change that is closing. “This is no longer about pilots,” he told the crowd. “This is no longer about proof of concepts.” 

Krishna organized the morning around what he called three forces converging to define who wins the AI era: the Al-first enterprise, hybrid cloud architecture and the quantum frontier.

Krishna kicked off his AI-first enterprise discussion by spotlighting Aramco, the world’s largest integrated energy company. IBM installed the company’s first computer back in 1947. Nearly 80 years later, Aramco SVP of Digital and Information Technology Sami Al-Ajmi was on stage to discuss what comes next.

Aramco has been building out its own AI stack for years, Al-Ajmi said, including models that predict rock formations and reduce drilling time, and a global optimizer that gives a real-time view of assets worldwide. Last year, he said, the company reported USD 5.2 billion in technology value, more than half of it generated by AI. “Eighty years ago, we used to buy machines from IBM,” Al-Ajmi said. “Today, we are collaborating to build the future of digital technologies and the energy sector.” 

On the product front, Krishna zoomed in on IBM Bob, a cartoon likeness of which appeared on stage in poster board form, eliciting laughs from the crowd. For those who aren’t familiar, Bob is an agentic platform that functions less like a coding assistant and more like a full member of a software development team, handling work from the earliest planning stages through shipping. Krishna noted that 80,000 IBM developers are already using it, with 45% average productivity gains. Samiya Kashif, Development Manager for IBM Software, demoed it on stage. “Bob reads the same handbook as we do,” said Kashif. “That’s how you make AI work at an enterprise scale. You give it accountability.”

To that end, Krishna also highlighted IBM Sovereign Core, which bakes policy enforcement into runtime, so enterprises aren’t caught off guard when regulations shift, and IBM Concert, a platform that uses AI agents to diagnose and act on system failures, rather than just surface them.

The hybrid cloud section was where Krishna got blunt about why AI initiatives fail. “The data is siloed,” he said. “The infrastructure is fragmented.” That, he argued, is where most AI initiatives actually stall—and it’s why IBM recently completed its acquisition of Confluent, the real-time streaming platform that 40% of the Fortune 500 runs on. AI agents are only as good as the data they can access in the moment, and Confluent puts live, governed data within reach across the enterprise.

Elevance Health CDIO Ratnakar Lavu also joined the stage to talk through what deploying AI looks like inside a heavily regulated industry. The health insurer handles 130 million calls annually across its member base and has already deployed AI-powered self-service tools—including a virtual assistant used by 22 million commercial members—to reduce that load. Lavu was candid about what it took to get there. “We started building AI solutions without actually really thinking through the process itself,” he said. “When you do that, you won’t actually see the efficiency or the right outcomes.”

Lavu’s advice to the room was to reimagine the process first, then build the AI solution around it. “Responsible AI should be embedded throughout that entire process—from ideation to building the solution to deploying it,” Lavu said.

Krishna closed on quantum by taking on the skeptics directly—the ones who dismiss it as science fiction and the ones who think the hype has gotten ahead of the reality. Quantum is the first genuinely new computing paradigm in 80 years, he argued, and it has crossed from a science problem to an engineering problem. “That’s not 20 years away. That’s not 10 years away. That’s within this year.”

Cleveland Clinic Chief EVP and Chief Research and Academic Officer Dr. Serpil Erzurum joined him to announce a breakthrough the two organizations reached together: a quantum-enabled workflow for simulating proteins—understanding a protein’s three-dimensional structure well enough to design something that changes it, which is how new drugs get made. “We can simulate a 12,000-atom molecule—to our knowledge, the largest to date,” she said. “It is a moment.” The practical stakes, she explained, are in drug discovery. “That’s therapy,” Erzurum said. “And that can make a difference in life.”

12:30 p.m. EDT | May 4, 2025

Ahead of IBM’s annual Think conference, IBM hosted Partner Plus Day at Boston’s Thomas M. Menino Convention & Exhibition Center. There, it became clear that the ecosystem multiplier effect is real. As IBM Senior VP of Ecosystem, Strategic Partners & Initiatives Kareem Yusuf, Ph.D., said in his keynote address to IBM partners on Monday, the focus is about the logical next steps: growth and ROI; or as he put it, “scaling this ecosystem in the most efficient and effective way to serve our clients and joint clients.” When a partner sells an IBM solution, deal value can grow manifold through services, upsell and cross-sell.

Yusuf was joined on stage by IBM Chairman, President and CEO Arvind Krishna, as well as leaders from several IBM partner companies, all backlit by a larger-than-life 3D luminescent IBM “Think” marquee.

“Clients are counting on us as they transform to AI-first,” Krishna said. “AI is moving quickly from just a source of cost and productivity gains to fundamentally changing business.”

Furthermore, the urgency for companies to use AI to transform business is accelerating: “infrastructure capex [capital expenditures] is approaching USD 1 trillion” by 2027, Krishna said. To this end, IBM is gearing up to prepare its clients for the new reality. Many companies, he noted, “have been stuck in a 1940s org structure”—that is, fragmented. “Here is finance, here is legal, here is sales, here is back office. We designed workflows that can move smoothly between these silos,” he said. Nothing inherently wrong with that, he observed, except that if you try to shoehorn AI into one silo at a time, you’ll only see marginal improvements, and you still have friction.

“The real advantage for AI adoption is that it allows you to think about the end to end, and then [ask], ‘How much of this can I make touchless, and how can I rethink my process?’” Krishna said. “Agents can cut across silos.”

Yusuf emphasized the importance of understanding “each of the diverse elements of [the IBM] ecosystem,” which he defined as sell partners, build partners and service partners.

A new IBM offering focusing on sell partners’ preparedness and ability to engage with clients, he unveiled, is the watsonx Workshop—an AI-enabled skill-building platform that allows for “coaching sessions, practice and even role play.” Another sell partner-centric innovation, he said, is Tier-One Support for the AWS Marketplace. Integrated directly into IBM toolchains, it allows sellers to link deals to the AWS Marketplace, thus meeting customers where they are, which will in turn “remove friction, drive volume, increase velocity and tap into committed spend.” 

Reyna Thompson, President, North America, TD SYNNEX, took to the stage to describe her company’s partnership with IBM. “One of the things that we heard from you is that renewals is really where there was the most friction,” she said. “So we pointed the API to the renewals effort and the team ... mapped the process end to end, and wherever we could point the API to fill a gap, we moved it to production. Everything else got automated.” The result? “85% of our renewals quotes are touchless. That’s a big deal.”

IBM build partner Nexar CEO Zach Greenberger said that one of IBM’s big value-adds to his company was bringing speed to market. “It’s one thing to collect the data,” he said, but IBM allowed Nexar to “take this incredible value data pipeline and actually build agents that we can deliver to customers.” He cited IBM’s Agent Catalog—a curated list of enterprise-grade applications from partners including Adobe, Box and Palo Alto. In particular, he called out the catalog’s “compliance, speed, legitimacy-in-market” and quick scaling as key benefits. He also lauded IBM Bob: “Everyone from engineers to product managers all the way to the finance team … is leveraging coding capabilities like Bob to help accelerate our business meaningfully.”

Vedavyas Avula, Founder and CEO of IBM service partner Pragma Edge, said that the most important thing IBM brought to the table for his company was trust: “Once we built the trust and delivered the right solutions for the customer, we were able to transform the way we are doing joint account planning.” This was invaluable, he said, in partner matching and sales coordination for Pragma.

Patrick Moorhead, Founder, CEO and Chief Analyst of Moor Insights & Strategy, related his experience with IBM in a fireside chat with Krishna. “First and foremost, IBM technology is hybrid, and not retrofit,” Moorhead said. “Second, IBM has a lot of depth and expertise in premier regulated industries.” And finally, “a lot of companies and a lot of programs have consulting, and a lot of them have software, but IBM connects the two.”