From Systems of Record to Systems of Action: Ron Davis on Agentic AI and the Future of ERP

As manufacturers face rising complexity, tighter margins, and growing pressure to act in real time, the role of ERP systems is undergoing a fundamental shift. No longer confined to reporting and record-keeping, ERP platforms are increasingly expected to participate actively in day-to-day operations—anticipating issues, guiding decisions, and helping teams execute faster.

In this interview, Ron Davis, Senior Vice President of Product Engineering at QAD | Redzone, shares how Agentic AI and Champion AI are transforming ERP from a passive system of record into a true system of action. He explains what makes Agentic AI fundamentally different from traditional machine-learning features, how it operates on the plant floor, and why this evolution is especially critical for mid-market manufacturers with lean teams. Ron also discusses governance, trust, and the practical realities of embedding AI directly into manufacturing workflows—delivering measurable value without adding complexity.

Q: QAD and AWS have introduced Champion AI as a transformative step for mid-market manufacturers. In practical terms, what differentiates Agentic AI from the traditional machine-learning features manufacturers have seen in ERP systems?

A: Traditional machine learning in ERP mostly analyzes data. It looks for patterns, makes predictions, and flags issues, but people still have to decide what to do and then go execute it.

Agentic AI is different because it’s built to act.

With Champion AI, agents understand the manufacturing process, know the goal they’re working toward, and can take or recommend actions across systems and teams.

That shift from insight to execution is what makes this meaningful for mid-market manufacturers with lean teams.

Q: How does Champion AI actually operate on the plant floor? Can you share examples of real workflows or decision cycles it autonomously executes today?

A: One example is our Line Lead Champion, operating within Redzone, our connected workforce solution for frontline teams. It helps line leads prepare for their shift by reviewing what’s already in progress, what’s coming next, and flagging risks—such as upcoming changeovers, predicted top losses, or known equipment and quality issues.

Using machine learning, it helps prepare for changeovers, anticipates where losses are most likely to occur, and surfaces previously successful actions taken in similar situations to help avoid the same problems. From there, it can recommend or initiate actions—like creating maintenance work orders, assigning follow-ups, or escalating issues early.

The result is simple: line leads start the shift prepared, instead of reacting after performance has already slipped.

Q: With Agentic AI now embedded across QAD Adaptive, what types of decisions can the ERP system now make proactively?

A: With Agentic AI built into QAD Adaptive, these Champions don’t wait for someone to click a button or run a report. They’re always paying attention to what’s happening and stepping in when it matters.

Productivity Champions take care of the routine, repetitive work in the background, freeing people up to focus on more important tasks.

Optimization Champions keep an eye on demand, supply, and constraints and proactively suggest, or kick off actions that drive real ROI, like freeing up working capital through better inventory decisions or simplifying how teams work with suppliers.

Implementation Champions help teams get up and running faster by guiding them through the process, automating setup and follow-ups, and removing friction along the way.

The end result is that the ERP starts helping teams get work done, not just keeping track of it.

Q: How does this shift impact core operational areas such as planning, maintenance, quality, and supply chain responsiveness?

A: The biggest impact is that these areas stop working off outdated information. When intelligence lives inside the workflow, planning adjusts faster because the system can point out the moment when the plan no longer matches what’s happening on the floor. Maintenance gets an earlier warning because the system can recognize patterns that typically lead to a problem instead of alerting after the fact.

Quality becomes more preventative – many issues start upstream, and an agent can surface those signals before they turn into defects. On the supply chain side, the system can immediately interpret what a disruption means for production instead of simply logging the delay.

The most transformational common change is speed. People get the right information at the right time, which lets them respond before small issues become larger ones.

Q: Many manufacturers still rely on ERP as a passive information repository. What does “system of action” mean in practice, and why is this transition so important now?

A: In our world, a system of action means ERP doesn’t stop at recording what happened. It continuously senses what’s going on in the process, understands the impact and prompts the next best step in real time. The system of record is still there as the foundation, but the agentic layer interprets those signals and drives action inside the workflow, not in a separate report.

That matters now because manufacturing runs faster than the old ERP model. If you only find out about a problem in tomorrow’s dashboard, you’ve already lost the chance to respond. A system of action is about catching that moment while you can still do something about it.

Q: How do you expect this shift to change the daily work of frontline teams and plant managers?

A: Frontline teams will spend less time hunting for what’s wrong and more time fixing what matters. Instead of digging through data or walking the floor to spot issues, the important signals show up in the tools they already use.

For plant managers, it means fewer “where did that come from?” surprises. They see risks and deviations as they form, not just in hindsight, so their day shifts from firefighting to steering. The system takes on the monitoring load so people can focus on judgment and leadership.

Q: Mid-market manufacturers often have limited IT resources. How does Champion AI balance advanced capabilities with usability for smaller teams?

A: We designed Champion AI so you don’t need a data science team to benefit from it. The agents are embedded in existing workflows for planning, scheduling, quality and so on, which means their functionality is right at your fingertips and teams don’t need to learn a new analytics environment.

Because it’s built on AWS and grounded in QAD’s manufacturing process knowledge, a lot of the complexity sits under the hood. What people see is a timely prompt, a recommendation or a “pay attention here” moment, not a model configuration screen. That’s how you give mid-market manufacturers serious AI without turning it into another IT project.

Q: Are there specific mid-market case studies or early results you can share?

A: We’re already seeing mid-market customers use Champion AI to bring issues forward earlier and stabilize their operations. Teams are learning about risks and drifts sooner, and they’re able to make decisions faster. 

Going back to that inventory optimization example, one early adopter of Champion AI was able to identify adjustments that translated to roughly $300,000 in working-capital savings. And with the support of our migration agent, customers are going live and seeing real value from QAD Adaptive with Champion AI in just 90 days – unheard of for ERP deployments. 

The biggest early win I see is time; time saved from digging through data, and time gained to respond before problems grow. That’s where the measurable value is showing up first.

Q: As AI becomes more embedded in operational decisions, how is QAD approaching transparency, explainability, and governance?

A: We believe that trust starts with context. Champion AI is grounded in the customer’s own data, processes and KPIs, not a generic model guessing from the outside. We make it clear what the agent is reacting to and keep humans firmly in the loop. The agent can recommend, but people decide.

On top of that, QAD has a strict AI policy around privacy, security and responsible use, and we’re building Champion AI on AWS infrastructure that’s already proven at scale in those areas. Governance is not an afterthought; it’s baked into how we design and deploy the agents.

Q: What safeguards ensure that Agentic AI actions are correct, compliant, and aligned with business rules?

A: The first safeguard is that agents operate inside your configured business rules and workflows. They learn your world – what you make, how you make it and what “good” looks like for you – and they stay within that frame.

The second safeguard is human oversight. We’re not pushing fully autonomous, black-box decisions into a factory. Champion AI proposes actions and changes based on clear signals; people review, approve and tune that behavior over time. That feedback loop keeps the system aligned with compliance, standards and how you actually want to run the plant.

Q: QAD has made one of the boldest moves toward operational AI this year. Where do you see the manufacturing technology landscape heading over the next 2–3 years?

A: I think we’re moving quickly away from systems that just report and toward systems that participate. Dashboards and static analytics won’t go away, but they won’t be enough. Manufacturers are going to expect software that understands their processes and can work alongside people in real time.

You’ll see more vertical depth, more role-based intelligence and a greater focus on the “messy reality” of how factories actually run. The winners will be the platforms that combine a strong system of record with an agentic layer that can sense, decide and respond in the flow of work.

Q: Do you view Agentic AI as the next major differentiator in ERP?

A: Definitely. The next big differentiator is whether your ERP can act, not just record. Agentic AI is the first serious step in that direction. When ERP carries agents that understand specific roles, care about the same KPIs as the people in those roles, and work with them in real time, that’s a very different value proposition than another reporting feature.

That’s where ERP starts to feel less like a database and more like a digital teammate.

Q: What is next for Champion AI? Are there upcoming capabilities or roadmap items you can preview?

A: The direction is to expand Champion AI across more roles and decision points in the factory. Today you see it in areas like inventory and core workflows; over time, you’ll see more agents tuned to specific responsibilities and pain points. We recently acquired Kavida.ai, a leader in Agentic AI for industrial companies, and expect that to significantly accelerate our product roadmap.

We’re also investing in making the experience feel more “ambient,” less like interacting with a separate AI tool and more like the system simply knowing when to step in and help. The long-term vision is that every critical role in manufacturing has a champion working alongside it.

Q: How will QAD continue supporting manufacturers transitioning toward AI-powered operational models?

A: We know not every manufacturer is starting from the same place, so we’ve built on-ramps like the Champion Benchmark to help them understand where they are and what’s possible. From there, we use approaches like Champion Pace and AI-assisted migration to reduce the pain of modernization and get to value faster.

Our job is to bring AI into their world without overwhelming them – keeping the intelligence inside the workflow, keeping people in control and keeping the focus on outcomes instead of hype.