Cloud adoption in manufacturing has long been seen as a game-changer, promising increased efficiency, scalability, and agility. Yet, for many enterprises, making the shift isn’t as simple as it sounds. Concerns over data integration, security, and long-standing reliance on on-premises systems continue to slow adoption.
In Malaysia, where manufacturing is a key economic driver, companies are increasingly looking at cloud-based ERP systems to stay competitive. Epicor, a provider of enterprise resource planning (ERP) solutions, works closely with manufacturers navigating these challenges. The company has been engaging with enterprises to discuss how cloud-based ERP systems can address these pain points and drive operational improvements.
In this interview, Frontier Enterprise speaks with Rich Murr, Chief Customer and Information Officer at Epicor. Murr shares insights on cloud adoption hurdles, the evolving role of AI in manufacturing, and the key risks companies should consider when transitioning to modular ERP systems.
What do you see as the biggest barriers to cloud adoption in Malaysia’s manufacturing sector, and how might enterprises overcome them?
I don’t think the barriers to cloud adoption are unique to Malaysia. A lot of manufacturers face the same challenges, especially those that have been leveraging on-prem systems for years. They have to overcome the same hurdles as any company moving from an on-prem solution to the cloud.
That can mean getting comfortable with the idea of paying a subscription instead of making a capital investment, trusting a third party to manage a significant part of their technology stack, and keeping pace with regular upgrades. But I don’t think these challenges are unique to manufacturing, and they’re certainly not unique to Malaysia.
There are companies in Malaysia that are already doing this well. There’s a track record for others to follow if they’re interested in the cloud. Really, it’s just a matter of shaking off some of that inertia — which companies absolutely can do. I’m here at Asia Connect 2025, and we have manufacturers who have already made that leap. They’ve shown that it can be done.
AI and automation are transforming manufacturing, but proving ROI remains a challenge. What common pitfalls hinder companies from effectively measuring the impact of these technologies?
That’s a timely question. One of the things we’re focused on is that, to really understand ROI, companies need to work on solving a specific business problem. They should look for opportunities where they can improve the speed of decision-making or automate a process using AI.
One challenge is that AI is such a broad category, such a broad topic that it can be hard to focus. Companies need to focus on solving very specific business challenges with the clear goal of driving ROI. For example, if you have many people working to solve a business problem, producing reports, or gathering data, you could say, “Let AI do that for us in a faster fashion.”
But you really have to focus. You have to go beyond just the excitement of generative chat and look at very specific business problems. If you do that, there are plenty of opportunities to identify significant returns.
What metrics should companies prioritise when evaluating AI’s impact?
To me, it’s time and labour. Are there things AI can do — sourcing data, generating reports, pulling analytics — that might take an end user much longer to complete? The answer is yes.
Can AI produce a meaningful data report that allows companies to make better business decisions? Can it do it faster than an individual? If businesses focus on areas where they struggle to gain insights or spend a lot of labour just to get those insights, that’s where AI should be deployed. And I think they’ll find plenty of opportunities.
Malaysia aims to move up the manufacturing value chain, but what structural hurdles could slow this transition? How can enterprise technology help?
I’ll answer the second part first. Enterprise technology can certainly help any organisation — if they leverage it effectively — to better understand and optimise their operations. For manufacturers, the more discipline and optimisation they have in their operations, the better suited they are for growth. A properly implemented ERP system plays a big role in that.
As for structural challenges — I won’t claim to be an expert on Malaysia’s manufacturing sector, but just being here in Kuala Lumpur, I feel like I’m learning a little. You look at what’s happening in Penang, for example, with medical device manufacturers, chip manufacturing, and industry partnerships — it’s clear that these efforts have already helped Malaysia move up the value chain.
So I think Malaysia has already demonstrated its ability to advance in manufacturing. It’s now a matter of building on those achievements.
Manufacturers today have more supply chain visibility than ever, thanks to enterprise technology. But turning all that data into action at scale isn’t always easy. What do you see as the biggest obstacles to making data-driven decisions operationally?
Good question — one we get a lot as customers look to tap into the full potential of their data.
I’d say the first challenge is making sure you have good data. To me, an ERP is 90% about data quality, insights, and analytics. If the data isn’t solid, the system won’t provide real value. And there’s no unique skill set required to get this right — it just comes down to good data stewardship, accountability, and ownership.
The second challenge is having people who understand the business and know how to leverage data effectively to craft insights. One reason companies struggle with this is because, honestly, it’s hard. You look at all the data inside an ERP — making sense of it to drive deep insights takes work.
This goes back to your earlier question about how companies can really leverage AI. Over time — and probably fairly quickly — AI is going to help businesses extract insights from their data more efficiently. Instead of relying on labour-intensive work to process and analyse data, AI will take on more of that workload. It’s going to mature quickly, and it’s absolutely one of the things we’re focused on as a company. We take a very practical approach to helping customers understand and leverage their data more effectively and more quickly.
More enterprises are shifting to modular, cloud-based ERP systems. What key risks do companies often overlook when making this transition?
There are a couple of things to consider. One is data. With a monolithic ERP, I do think the data challenge is a little easier, but when business functions are crossing multiple ERPs, you’re moving data between applications, and you’ve got to do that well. Integrating data effectively is a skill set that requires careful thought and good execution, so that’s something to keep in mind.
Now, the advantage of modular ERPs is that they are purpose-built and may have deeper expertise in the functions they support. But if you’re crossing ERPs, you’ve got to think about user training as well. Employees may need to work across multiple applications, which adds complexity.
And of course, you have to consider security, performance, and tuning. Just because you optimise one ERP for performance doesn’t mean it will work seamlessly with others. In fact, you could end up creating issues downstream with another ERP system. So when companies go for a best-of-breed approach, there’s quite a bit to think through.
Looking ahead, what technology shifts do you think will have the biggest impact on manufacturing over the next five years?
AI is going to play a huge role. We’re still in the hype cycle, and it’s very early in its development, but we’re just getting started, and the opportunities are tremendous.
Over the next six to 12 months, we’ll see companies bringing the first truly practical iterations of AI to market. But it’s going to go far beyond that. The innovations are just beginning. It’s hard to predict exactly what’s coming, but we’re going to see enormous change and real benefits for companies that figure out how to leverage AI effectively. That said, AI isn’t necessarily going to make everything easier overnight — there’s going to be some lift required to fully tap into its potential.
It’s not just generative AI, right?
That’s right. Generative AI is what people get excited about, and it’s fascinating to see what it can do. But AI has been around for a long time. We’re going to see generative AI combined with machine learning and other AI technologies, making it much more accessible — even to non-technologists.
It’s going to make complex systems easier to engage with and leverage. But yeah, I do think AI is the focus area for at least the next 36 months, which is a long time in the tech space.
