Dominic Jainy is a seasoned IT professional who has spent years navigating the complex intersection of artificial intelligence, machine learning, and the rigid world of legacy infrastructure. With a background that spans the evolution of blockchain and the rapid rise of generative models, he specializes in helping organizations bridge the gap between “the next big thing” and the practical reality of how a business actually functions. Our conversation focuses on the tension between the high-speed allure of AI-first strategies and the often-neglected foundation of Enterprise Resource Planning (ERP) systems. We explore how to identify genuine AI opportunities, why messy data is the ultimate project killer, and the strategic importance of using intentional friction to build systems that last.
Many organizations find ERP modernization slow and disruptive compared to the fast pace of AI tools. Why does this “heavy” feeling often lead companies to prioritize AI even when their foundation is shaky?
The reality is that AI feels like pure momentum—it is shiny, fast, and promises an immediate impact that creates a sense of excitement in the boardroom. When a company rolls out a tool like Microsoft Copilot, they see instant feedback, whereas ERP work feels heavy because it forces leadership to look closely at the “technical debt” and the uncomfortable ways a business actually runs under the hood. I have seen countless teams gravitate toward AI because it feels like a shortcut, but if the underlying ERP is inconsistent or hard to trust, that AI tool only serves to amplify the existing mess. It is easy to underestimate the risk at the start, but without a solid foundation, you are essentially building a high-speed engine on a frame that is already rusted through.
While the risks are clear, there are instances where an AI-first approach provides real value. In what specific scenarios should a company lead with AI before completing a total ERP overhaul?
Leading with AI makes a lot of sense when you focus on “quick wins” that can automate repetitive, stable tasks like data intake, routine reporting, or simple approval workflows. These initiatives create visible momentum and improve the daily user experience by adding natural language interfaces that make clunky, older systems easier to interact with. More importantly, AI can act as a diagnostic tool; by running AI-driven analytics, you can uncover patterns and trends that legacy platforms were never designed to catch, effectively highlighting the gaps in your current data. This serves as a way to “test” your future-state design, showing you exactly where an ERP improvement will have the most significant impact before you commit to a full system replacement.
We often hear that data is the fuel for AI, but how does a fragmented or overly customized ERP environment specifically cause an AI strategy to backfire?
The most common failure I see is the “poor data equals poor outcomes” trap, where an AI initiative magnifies errors across the organization at a scale a human never could. If your ERP environment is a patchwork of fragmented systems, your AI models will produce results that are impossible to trust, leading to misplaced expectations and a loss of confidence from stakeholders. Many of these projects stall after a successful pilot because the core processes simply aren’t mature enough to support a broader, enterprise-wide adoption. You end up with a team that is overwhelmed and fatigued, trying to manage sophisticated AI outputs on top of outdated tools that are already a struggle to maintain.
If an organization is ready to move forward, how should they conduct a readiness assessment to ensure their AI and ERP strategies are actually aligned?
A truly effective readiness assessment requires a partner who is willing to be honest about the five key pillars: process maturity, data quality, user sentiment, leadership alignment, and technical debt. You have to take a hard look at whether your data structure is actually connected or if it is sitting in isolated silos that will starve an AI of the context it needs to be useful. It is about asking if the leadership is aligned on long-term goals or just chasing a headline, and whether the users are actually ready to adopt new workflows or if they are already reaching a breaking point with existing systems. Working through these questions gives you a clear, honest picture of whether you can leap into AI today or if you need to spend six months tightening the ERP foundation to save yourself years of frustration later.
You’ve mentioned that mapping out the “current and future state” is vital. How does this exercise reduce the internal resistance that usually accompanies large-scale technological changes?
When you take the time to map out how work actually happens today and compare it to where the business needs to go, you take the guesswork out of the transformation process. Clarity is a powerful antidote to resistance; when people understand exactly why a change is happening and how the new AI-enhanced ERP will make their specific job easier, they are much more likely to buy in. This exercise allows teams to see the connection between the tools they use and the outcomes they produce, making the transition feel less like something forced upon them and more like a necessary evolution. It reduces the “fear of the unknown” by providing a transparent roadmap that respects the current workflow while clearly defining the benefits of the future state.
Interestingly, you suggest that AI pilots should be designed to “expose limitations” rather than avoid them. Why is “intentional friction” more valuable than a perfectly smooth pilot program?
It is a common mistake to design a pilot so perfectly that everything runs smoothly in a controlled environment, only for the system to collapse when it hits the “messy” reality of daily operations. I advocate for letting the friction show up early—let the pilot expose where the data is inconsistent, where the processes get tangled, and where the ERP starts to strain under the new demands. Those moments of failure are incredibly valuable because they point directly to the gaps that need your attention before you scale. If you don’t find the breaking points during the pilot, they will find you during the full rollout when the stakes are much higher and the cost of correction is ten times more expensive.
Change management is often treated as a secondary concern, but you argue it’s the heart of the project. What are the practical steps to building trust with users who might be skeptical of AI?
Successful transformation is always more about the people than the technology, and that starts with bringing users into the design process on day one, not after the decisions have been finalized. You have to create a genuine space for feedback—and then actually act on it—so that the staff feels like they are co-authors of the new system rather than victims of it. Highlighting small, tangible wins along the way is essential for making progress feel real and achievable, which helps build the trust necessary to scale the solution later. Without that foundational trust, even the most technically perfect AI-integrated ERP will struggle to stick because the people using it will find ways to bypass the system.
Rather than a massive “all-at-once” overhaul, you advocate for thinking in phases. How does a phased approach allow the ERP and AI to mature together over time?
Modern ERP should not be viewed as a one-time, massive replacement; it works far better as an evolving system that you build on in strategic stages as your business grows. By thinking in phases, you can add AI capabilities gradually as your data becomes cleaner and your processes become more disciplined, ensuring that each new tool is grounded in a solid reality. This continuous progression allows your team to adapt to new technologies at a sustainable pace, preventing the burnout that often follows “big bang” implementations. Done correctly, your AI doesn’t sit on the sidelines; it grows alongside your ERP, becoming smarter and more integrated as your overall digital maturity increases.
What is your forecast for AI-integrated ERP systems?
I forecast that the era of the “monolithic” ERP is ending, replaced by a “living” ecosystem where AI and ERP are treated as complementary parts of a single, fluid transformation. In the next few years, we will stop seeing AI as a separate layer and start seeing it as the primary interface for all enterprise data, but only for those organizations that did the hard work of cleaning up their data today. The companies that will thrive are the ones that balance innovation with stability, moving quickly where it makes sense but having the discipline to slow down and fix their core systems when it matters. Ultimately, the goal is to create a path where your technology gets smarter every day without leaving your people or your foundational stability behind.