“Tons and tons of money is being poured into AI, but we’re not seeing the results. Strategy is great, spending money is great, but without execution, it’s completely pointless,” said Mike Capone, CEO of Qlik, during his keynote speech at the Qlik Connect 2025. The annual summit held by the software company has industry leaders gathered to throw light on the current state of AI implementation across enterprises, underscoring a concerning reality that amid massive investments most organisations are grappling to move from strategy to execution.
In his keynote address, Capone spoke about the state of AI deployment. “While 86 per cent added AI strategy, only 26 per cent actually deployed at scale. It’s a massive gap. It’s tragic,” Capone noted. This gap in implementation is seen as a critical challenge for companies that are rushing to embrace the AI wave, especially the recent phenomenon of Agentic AI that focuses on autonomous, goal-directed systems built to take independent actions.
“80% of companies are saying they’re investing in agentic AI,” Capone explained. “But again, here’s another gap. Only 12% of companies say or feel that their data is actually ready for agentic AI. So here we go again. Let’s everybody jump in the water on agentic, but forget that you got to do the work.”
The data foundation challenge
Moments into the keynote, Capone welcomed Ritu Jyoti, group vice president and general manager of AI at IDC. On being asked what are the reasons that are impeding AI adoption, “the first factor is that they (companies) have fragmented data. They have disjointed data and disparate systems leading to ineffective use for AI.”
The IDC executive added that the other barriers include AI strategies operating in isolation from corporate strategy, a lack of AI-ready workforce, and overall a cultural resistance due to job security concerns.
Jyoti emphasised the transformative potential of agentic AI while separating it from generative AI: “Gen AI was all about augmenting a human and improving productivity and operational efficiency… If you think about agentic AI, the focus is on agility, adaptability, and timing,” she explained.
Jyoti later illustrated her POV with an example. “If you have a fully autonomous software engineer, you simply provide a high-level goal, like ‘build an app to manage company-wide logistics and inventory’, and the system delivers it. With generative AI, progress happened at a human scale, improving individual productivity. But with agentic AI, it’s humans plus digital labour, allowing work at a speed, scale, and precision no human-only team can match.”
Story continues below this ad
Implementation in real world
Tom Mazzaferro, chief data AI and Analytics officer at Truist also made a brief appearance at the keynote. Mazzaferro shared insights on how the US bank is navigating the current challenges. “Every big bank still operates in a hybrid ecosystem, both on ground, in the cloud. For us, it’s all about how you bring it together to deliver our business strategy and service our customers and our clients as we go forward,” he explained.
Mazzaferro stressed on the importance of partnerships in successful AI implementation. “For us, we can’t do it alone. We need to rely on partners, on key SMEs, on key solutions to deliver success for both our clients, but also for our teammates, for our employees.”
When asked what advice he would give to others embarking on similar journeys, Mazzaferro recommended, “figure out what you’re going to start. What do you want to achieve? How are you helping your business achieve their goals, their success, their outcomes? How is the technology that you’re building and enabling helping your clients be successful?”
Regardless of the challenges, organisations are thriving. “The early adopters, they are kind of being mindful of the risks. They are being careful about their autonomy… but they’re jumping ahead and failing fast, learning their lessons, setting up their structure,” Jyoti explained. She cited Johnson & Johnson as an example, which according to The Wall Street Journal is “using AI for chemical synthesis during drug discovery, to accelerate drug discovery.
Looking ahead
Capone highlighted that the sole way to succeed was by focussing on the foundation that is trusted data, rather than pursuing specific AI models. “It is not about models at all. It’s about harnessing your data. It’s about trusting your data, and it is about embedding AI where it drives outcomes that are tangible, measurable, real.”
Story continues below this ad
Both Capone and Jyoti addressed concerns about AI replacing jobs. When asked if everyone will lose their jobs to AI, Capone responded, “I don’t think so. I think people will lose their jobs because somebody got better at AI than you, not because AI took your job.”
On the other hand, Jyoti shared a personal anecdote.”I always joke with my son when I did my engineering, I didn’t get a chance to use a calculator. But when he did his engineering, he used a calculator. That doesn’t make him less intelligent than me.”
Capone wrapped up his keynote with a call to action focused on execution. “The race isn’t coming. It’s already on. The winners are not leaning. They’re executing now… You’ve never been more important to your organisations than you are today. So the question is, are you moving fast enough not just to keep up, but to win?”
The author is attending Qlik Connect 2025 in Orlando, US, at the company’s invitation.