Enterprise resource planning, or ERP, was designed for predictability — but today’s operational landscape is anything but familiar. As supply chain disruptions become the norm rather than the exception, the static nature of the traditional mid-market manufacturing ERP is no longer sufficient for agile enterprises.

To address this, QAD Inc. is focusing on tangible ERP returns by strategically pivoting from providing a passive system of record to a dynamic “system of action,” according to Amit Sharma (pictured), president of manufacturing ERP at QAD. This system is designed to interpret data and suggest solutions in real time, leveraging artificial intelligence where it makes the most sense for return on investment.

QAD’s Amit Sharma talks with theCUBE about how agents are transforming manufacturing ERP into a proactive system of action on the factory floor.

“We are not [adding] AI bolt-on bells and whistles,” Sharma said. “We are focusing on only those use cases which deliver measurable and meaningful ROI.”

Sharma spoke with Scott Hebner, principal analyst at theCUBE Research, and Paul Gillin, enterprise editor at SiliconANGLE, at the QAD Champions of Manufacturing event, for an exclusive interview on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed the evolution of ERP systems and the integration of agentic AI into factory floor operations. (* Disclosure below.)

Modernizing mid-market manufacturing ERP

For the past 40 years, ERP systems have primarily functioned as digital ledgers, recording transactions to ensure books were balanced and inventories replenished, Sharma told theCUBE. However, in a modern manufacturing environment, recording a problem without triggering a solution is insufficient.

“If a supplier sends a note saying that this entire delivery is going to be delayed, that information should not just sit in ERP,” Sharma said. “That’s an action, which means now what we have committed to our customers, we can’t deliver it. … This is where [the] system takes action based on the events that are happening.”

To facilitate this shift, QAD is deploying what it calls “Champion AI” agents. These are specialized digital assistants integrated into specific workflows — such as procurement or quality inspection — that share the same key performance indicators as their human counterparts. The goal is proactive problem-solving rather than reactive reporting, Sharma explained.

“We want to apply AI to have those Champion agents which carry the same ambition, same objectives, same [key performance indicators] to their human counterpart,” Sharma said. “The idea is, for example, if you are a procurement manager and you feel there’s a risk … AI should be able to do this before you and tell you the alternative plan.”

Speed is critical for mid-market manufacturers, many of whom operate on thin margins and cannot afford lengthy, disruptive software overhauls. QAD has introduced an ambitious 90-day implementation model for greenfield sites, utilizing AI to handle data migration and templated workflows based on industry best practices, Sharma noted.

“We have real customers who are taking that journey with us as we speak,” Sharma said. “We are doing it 90 days in reality, not on the [PowerPoint presentation].”

By focusing exclusively on the manufacturing vertical rather than diluting resources across dozens of industries, QAD aims to provide a depth of functionality that generalist platforms struggle to match, according to Sharma. This allows for a tighter integration of AI directly into the “system of action,” effectively amplifying human capability on the shop floor.

“You can call that information worker 2.0,” Sharma said. “It’s going to augment humans and amplify their potential.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of the QAD Champions of Manufacturing event:

(* Disclosure: TheCUBE is a paid media partner for the QAD Champions of Manufacturing event. Neither QAD, the sponsor of theCUBE’s event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE