Copyright Forbes
Keeping up with the conversation on AI is a challenge on par with deploying it with a positive ROI. Every week brings a new contradiction, whether it’s AI is making hiring redundant while making hiring more essential or promising to reduce friction while simultaneously adding new layers of complexity. It is both the most overhyped and the most transformative technology of our time, and which version we get is determined entirely by the mindset we bring with us. Making sense of its real impact requires slowing down and reading more than just the tea leaves everyone is busy interpreting or MIT reports on the mileage others are getting from the technology today. This time, perspective comes not from rising above the noise, but from digging beneath it. While the headlines chase novelty and provocation, the most profound transformations are happening somewhere deeply insulated from the noise. By far the most interesting AI-driven revolution is happening inside the core systems that have powered enterprise work for decades. The software that runs the world’s payrolls, supply chains, and projects is being reimagined from the ground up with none of the fanfare the shiny VC backed point-solutions are receiving. And yet, these are exactly the stories worth paying attention as we search for the signal amidst the noise. One example lies inside a forty-two-year-old company embedded in how the enterprise world operates at scale. Under CEO Bob Hughes, Deltek is quietly leading the shift from systems of record to systems of reason, and this is a lesson in what happens when the systems that run our work begin to think for themselves. MORE FOR YOU When systems start thinking The last revolution in enterprise software was built on the back of rows and relationships between them. Databases became the nervous system of business, and dashboards its eyes. In the early years, Oracle and SAP defined how work was recorded. Then Marc Benioff left Oracle to build Salesforce, and cloud-based systems changed how software was bought and sold. That era was about bringing data together and making workflows visible. But it remained a world of systems that captured what happened, not one that understood why. Enterprise applications have always lived deep in the body of the organization and they remain the steady heartbeat that keeps the lights on to the point that replacing them feels like open-heart surgery that most CTOs wouldn’t even dare to suggest. Yet the pressure to modernize is building from every direction, and somewhere in ERP land, the boardroom itself is sweating because the heat is on to make these systems reason, not just record. Bob Hughes, CEO of Deltek, understands that pressure better than most. “We have to move from systems that look backward to systems that anticipate what is next,” he told me. “That is the promise of AI, and it is what our customers are asking for.” Deltek’s customers sit at the center of the world’s most consequential workflows from defense, construction, engineering, to government. None of these are known as fast-moving sectors, but they are decisive ones. Their projects are too critical to fail, their compliance too strict to bend. For decades, they have valued reliability above reinvention, but now with AI, they are suddenly at the forefront of how legacy ERP systems are finally buckling under the pressure. “Our customers are smart, and their demands are rising,” Hughes said. “They want AI that is useful, not flashy. They need a platform that unlocks the power of AI while ensuring the governance and compliance their industries require.” More than anything, these clients are demanding intelligence from the systems they run on. Yesterday’s innovation is today’s baseline, and a reporting dashboard is no longer impressive. Even AI itself is no longer a selling point, let alone a unique one. An AI that can reason, recommend, and automate is the new expectation. Hughes said that meeting those expectations required rethinking the company’s architecture from the inside out. “We have spent the last two years rebuilding how our platform connects projects across their entire lifecycle,” he explained. “That means using AI to turn data into foresight, to help customers plan better, execute faster, and make decisions with confidence.” At Deltek ProjectCon this week the company just introduced the first wave of that redesign. The company unveiled three core capabilities under its new intelligent platform: AI-generated proposals, enterprise risk management, and an agentic financial close system. Together, they are meant to bring reasoning into the daily rhythm of work, and they are a bellwether of what the core-application industry will be undergoing at large soon enough. The shift of approach is subtle but seismic in its long-term impacts. The innovators’ dilemma Christensen once enunciated is no longer whether to cannibalize existing margins in order to innovate, but how to meet dynamic client expectations without breaking what already works. For Hughes, that means embedding reasoning into the system itself. And that involves teaching the system to think through the workflow. “AI is still early, but the potential is undeniable,” he said. “Our customers need an intelligent platform so they can move with confidence throughout the project lifecycle. No one else empowers project-based businesses like we do.” Given how the industry is evolving, the winners will not be those who move the fastest, but those who bring intelligence into the core, not just the applications around it. When the core starts to open As reasoning seeps into the heart of these systems, the definition of an enterprise application begins to change. The interface no longer ends at a dashboard. It extends into decisions, predictions, and self-correcting workflows. The core is learning to think, and in that process, it is also beginning to open up in ways it hasn’t before. Raj Koneru, the CEO of Kore.ai, knows this transformation well. His company has been at the forefront of enterprise AI for more than a decade, and his vantage point offers a panoramic view of what is coming. “Legacy enterprise apps are facing a death by a thousand cuts,” he told me. “New tools are appearing everywhere, expectations are rising, and AI is seeping into every corner of the workflow. The core that once felt untouchable is now more open than ever.” Koneru believes that what used to be called the “application layer” is where the real change is happening. “We have moved from understanding to execution. The context window is expanding, the orchestration is improving, and we are moving closer to semi-autonomous systems,” he said. “The goal is not to replace humans but to bundle the work intelligently so that we know what can be automated, what should stay human, and how those pieces connect.” That bundling will define the next decade of software, and soon enough we may see the prompt bar will replace the menu. The enterprise app of the future may look less like a dashboard and more like a conversation. Deltek’s approach aligns with that vision. Its AI orchestrator, Dela, connects the entire project lifecycle from winning, planning, executing to analyzing through agentic reasoning. The idea is not to add a chatbot on top of the system, but to make the system itself capable of dialogue with the humans that use it. Hughes described it plainly. “We are not building AI as a separate experience,” he said. “We are embedding intelligence into the way people already work. It should not feel like a different system. It should feel like the same work done more intelligently.” This is where most AI transformations fail. It is easy to bolt on an API and call it progress. It is far harder to rebuild the underlying logic of a system so that intelligence becomes a native function rather than a foreign add-on. As Hughes put it, “Deploying AI requires rethinking how processes flow, how data connects, and how humans interact with both. You cannot automate your way to insight. You have to design for it.” That philosophy separates companies that are experimenting with AI from those that are internalizing it. It is the difference between having a system that assists and a system that reasons. A new architecture of enterprise intelligence The move from systems of record to systems of reason is not a single leap. It is an architectural evolution. The databases of the past stored information. The platforms of the future will understand intention. The real innovation is not in adding AI to what we already have, but in reimagining what the core of an organization can be when its systems begin to think. When every project, every process, and every ledger can learn from its own history. This evolution will not be painless, just like enterprise reinvention never is. But the direction is unmistakable. As Koneru said, “We are somewhere between semi-autonomous and fully autonomous. The challenge now is to build the confidence and the guardrails to move responsibly.” The companies that will lead the next era of enterprise technology will be the ones that treat AI not as a feature, but as an approach to serving clients with results, not product specs. That is what Deltek is betting on. By redesigning its architecture to embed reasoning, not just automation, it is showing what the next generation of enterprise systems will look like. When that becomes the norm, the core of the company will no longer be a static record of the past. It will be a living system of reason, capable of improving itself over time. And that, quietly and without fanfare, is the real AI revolution already underway.
