Manufacturing is in the middle of a once-in-a-generation reset. Tariffs, reshoring and legislation like the CHIPS and Science Act are restructuring supply chains and driving billions in new investment. 

Across the U.S., new plants are being built. But the pressure is just as real as the opportunity. Volatile demand, persistent labor shortages and a global market that favors whoever moves fastest.

In a recent Gartner piece, QAD | Redzone CEO Sanjay Brahmawar explored why this moment demands a new kind of ERP. Gartner’s Greg Leiter and Denis Torii echoed similar sentiments in the research note, Innovation Insight: AI Is on the Cusp of Reshaping ERP. Combined, the articles are a powerful indicator that the systems manufacturers have long relied on are fundamentally unprepared for the complexity they now face. 

The solution isn’t a better version of the same tool. It’s a new operating model entirely.

The Problem with Traditional ERP for Modern Manufacturing

ERP was never designed to make decisions; it was designed to track them. A transaction happens, the system records it. An inventory level drops, someone notices and reacts. A production schedule changes, someone rebuilds it manually. Every step involves a person in the loop, which was fine when manufacturing moved at a certain speed. It’s less fine now.

“Traditional ERP was built for structure and control, not for the pace and complexity manufacturers face today. These systems track data but don’t act on it,” shared Sanjay.

The problem isn’t a lack of investment or effort. It’s simply that legacy ERP systems, especially those still running on-premise, weren’t built to sense changes in real time or reason through options. The result is slower decision-making, reliance on manual processes and a disconnect between the speed of the market and the speed of the business.

What Makes Agentic AI Different?

Most manufacturers are familiar with AI and using it in their operations to some degree – with mixed results. A great example of this, according to Gartner’s Greg Leiter and Denis Torii, is generative AI. There are some real use cases, such as supplier communications and drafting job posts, but the technology is fundamentally reactive. 

Agentic AI is different because an AI agent doesn’t wait to be asked. It perceives what’s happening, reasons toward an objective and acts, escalating to a human where appropriate. Leiter and Torii describe this as a spectrum from low to high agency, and we’re still in the earlier stages. That said, the gap between where Agentic AI is today and where a fully autonomous system would be is closing faster than most enterprise software has historically moved.

And that’s significant, explained Sanjay, because agentic AI is so much more than a layer of intelligence. It’s an operating model for how ERP works. Embedded AI agents can detect anomalies, simulate outcomes, recommend actions and even execute next-best steps autonomously, escalating only when needed.

The Value of Agentic AI on the Factory Floor

The power of agentic AI is visible in several critical aspects of manufacturing operations. Supply chain management is one of the most immediate. AI agents can identify unusual supplier transactions and manage the initial resolution steps independently, empowering procurement leaders to focus on other important tasks.

Agentic AI is also a game changer on the labor side, which is a massive opportunity considering the industry is facing a 3.8 million skills gap by 2033. It can help with simple tasks like automating job posts and candidate sourcing, but also address a much larger challenge: preparing the next generation of manufacturing workers. Younger workers demand technology that feels as intuitive as the apps and devices they interface with every day, making outdated systems a recruiting liability. Modern, conversational AI interfaces are essential.

On the plant floor, agentic AI can model multiple production scenarios, create new schedules in response to shifting demand and surface actionable insights from data that would otherwise sit in a dashboard waiting for human review. In an environment where keeping up with delivery demands and shifting customer expectations are key, agentic AI provides the competitive edge that modern manufacturers need. 

For new facilities, the technology is especially transformational. Manufacturers using QAD’s Agentic AI capabilities have gone live in 90 days. Traditional ERP implementations at greenfield sites have routinely taken six to eighteen months.

Gartner’s Hot Take on Agentic AI in ERP

The market data makes the future of agentic AI clear. In 2024, 14% of ERP spending went to applications with GenAI or agentic AI capabilities. Gartner forecasts that number to reach 62% by 2027. That’s not gradual adoption – it’s a market undergoing massive change in a three-year window. 

Gartner’s Greg Leiter and Denis Torii also had some good advice for manufacturers considering agentic AI: keep an eye out for “agent-washing” – the practice of relabeling chatbots or basic automation as AI agents. Instead, anchor every vendor evaluation to a specific, proven use case. And don’t forget that quality is king. Agentic AI with weak data governance doesn’t just underperform. It can compound bad decisions faster than a human would. 

For most manufacturers, the practical question is where to start. Gartner recommends identifying the specific processes that cost the most in delay or manual effort, then mapping them to available AI use cases with proven outcomes. 

Agentic AI in ERP, like ChampionAI, is no longer theoretical. The competitive gap between manufacturers using traditional ERP and intelligent systems like Adaptive ERP will continue to widen. The real question is whether to move now or wait for competitors to make the move first.