Oracle - ‘Generative AI will allow supply chain management to shift focus from details to exceptions’

Derek du Preez Profile picture for user ddpreez March 19, 2024
Jon Chorley, Oracle’s Group VP of Product Strategy for SCM, says that Generative AI will change the way that we think about managing supply chains.

A birds eye view image of a number of trucks and vehicles in an industrial car park
(Image by marcinjozwiak from Pixabay )

Oracle last week announced a swathe of new Generative AI capabilities within its Fusion Cloud Applications Suite, embedded within existing business workflows across finance, supply chain, HR, sales, marketing and service. Generative AI is making the adoption of cloud-based applications even more compelling, given the opportunity to drive efficiencies and allow workers to do more ‘value added’ work (according to its proponents). 

Each ‘function’ within the enterprise will likely have different applications of Generative AI - depending on where the use cases are most compelling for early adoption. Supply Chain Management is notoriously one of the most challenging, complex and fragmented technology areas for buyers, where they have historically relied on best of breed solutions, integrated together, in an attempt to get visibility across areas such as manufacturing, planning, procurement, warehouse management, transportation etc. 

The advancement of cloud adoption has, however, helped promote the idea of a SCM ‘suite’, which in theory allows enterprises to get better visibility into their supply chains, using cross-platform data. Often though that is far from the reality, given that supply chains rely on data from other organizations and are often impacted by external factors. For instance, the past few years have seen huge shifts in supply chain strategy for enterprises, as they have faced inflationary pressures and been pushed to reconsider suppliers as a result of wars in Ukraine and the Middle East. 

With all of this in mind, we got the chance to sit down with Jon Chorley, Oracle’s Group VP of Product Strategy for SCM, to discuss how priorities are changing for buyers as they consider their supply chain needs, but also how the introduction of Generative AI tooling may force a change in how they think about SCM as a whole. 

Chorley said that there is a recognition amongst enterprise supply chain professionals that “things won’t ever go back to normal” and that this is prompting organizations to rethink their technology needs. He said: 

There won't ever be another normal. We're in a period of continual change and disruption. And that really has a lot of impact on the supply chain. There’s a re-appreciation of risk. I think the more you optimize, the more you add risk.

I think people are recognizing that, analyzing that more effectively. They’re looking at ways that they can look at alternate sources of supply, maybe near-shore. They’re really looking at other hedging strategies as well. 

The headache of just dealing with inflation and how that's impacted the cost of inventory carry…people have reacted to shortages and then have had to deal with the lump of inventory, managing all of that. This has just really shone a light on some of the weaknesses of people's existing supply chain solutions. 

There is an increased need for flexibility from buyers assessing their supply chain tools, given the amount of changes to strategy that have been forced in recent years. However, having better visibility is also needed, as organizations seek to understand their supply chains more clearly and reduce the impact of any potential changes/risks. 

This emphasis on change, as buyers respond to external changes, is forcing organizations to consider not only cloud-based tools, but cloud-based application suites for SCM. This is because on-premise applications are less agile and require a commitment to a ‘plan’ over extended periods of time. Chorley explained: 

That's the nature of supply chain. It's always about trade offs.

Cloud solutions are built for change in ways that on premise systems are not. Every quarter we deliver around 200 new capabilities across supply chain. That is a platform designed to support change. And then, also, the flexibility that we have to build into a single platform to support all of our different customers means that any one customer has a lot of flexibility in how they choose to operate. 

If you're shifting into a new business model, selling services instead of selling products, you can do that. You have this complicated, ever changing environment. And we have a platform that can better support that. 

A digital thread

Chorley argued that supply chain management is largely about ‘inter-process communication’, which is better suited to a suite-based approach in the cloud. On premise applications, integrated in an attempt to gain visibility, is an approach that is prone to error - and also makes delivering change difficult (things break easily!). He said: 

You may have a bell or whistle in a given feature that you like from another company, but on balance, that pales in comparison to having a solution that's designed to work together.

There’s a recognition that companies want to have a complete digital thread. If you have a planning solution, but it's not well integrated - well, you may have a good plan but it breaks down when you try and drive the change through the supply chain. 

Or if you have a good PLM solution, but it’s not integrated into your manufacturing process, then it breaks down when you try and drive that into actual production. And so you can see the challenges of having an individual best of breed strategy. 

Even having a well integrated cloud suite has its challenges, he added: 

People are going to the cloud because they don't want to own the maintenance and management of the software. But if you have five or six different cloud solutions, you own the integration between those products. And so you've really eroded one of the core values of cloud. It's not that we do everything for everybody, but we do most things for everyone. And that's definitely a better place to be.

Again, if you have multiple different systems, your ability to analyze that holistically is much reduced. It's much more complicated. You probably have multiple definitions, multiple definitions of supplier, and so on. Those things become much more difficult to report accurately against. Can you do it? Yes, you can, but it's a heavy lift that the individual company has to do.

The possibility of Generative AI

The AI capabilities announced for the Oracle Fusion Cloud Applications Suite, for SCM specifically, include: 

  • Item description generation - Generative AI will help product specialists quickly generate standardized product descriptions that highlight SEO keywords. Using Oracle Product Lifecycle Management, the vendor hopes to help organizations save time, reduce errors, and improve the overall quality of product descriptions to increase customer engagement and boost sales.

  • Supplier recommendations - procurement professionals will be able to quickly add suppliers to their organization’s supply chain, according to Oracle. With generative AI, supplier recommendations will be embedded in Oracle Procurement and organizations can use information such as product descriptions, purchase categories to identify suppliers, improve sourcing efficiency, help lower costs, and reduce supplier risk.

  • Negotiation summaries - with a generative AI-powered assisted authoring tool, embedded in Oracle Procurement, Oracle hopes to help organizations accelerate negotiations, increase savings, reduce risk, and maximize supplier outcomes.

Chorley was keen to note that many Generative AI capabilities are available now within its Fusions applications to augment the supply chain management processes - this isn’t future gazing. On the types of areas that the vendor is focusing on, he said:  

We’ve initially focused on assisted authoring use cases, summarization use cases, and then also selection type use cases. So, for example, in an RFQ you could ask: ‘please advise me sensible suppliers to send the RFQ to that may not be my normal suppliers’. 

We can use AI to better predict probable demand patterns of new products, for which you don't have demand history, by looking at related products and constructing a demand curve for them. All of those capabilities are there. 

However, Oracle is still just ‘scratching the surface’, according to Chorley. He believes that Generative AI is going to be a transformative technology, where users aren’t having to struggle in the weeds - but can instead take a big picture approach to managing their supply chains. Chorley explained: 

It’s a movement away from dealing with details, to dealing with well informed exceptions and advice, which help you address those exceptions. I think that's going to be the real change. 

And also, frankly, a reduction in the need for highly specialized skills to do this stuff. Even in the assisted authoring area, I think there's some great use cases. A typical shift supervisor would have a morning meeting and we could make sure he has a completely auto generated briefing for what happened the day before. All of that can be enormously useful in driving efficiency.

In terms of advice for buyers that are looking at the advancements in Generative AI and thinking about the impact on their supply chain management choices, Chorley argues that they too should keep a focus on the big picture. Lots of choice is going to become available, but buyers should try to have a long-term view of what outcomes they desire. He said: 

Where do you want to be in five years? I think if you take a lot of tactical decisions, you often end up with: how did I get to where I am? 

You've perhaps made some independent tactical decisions that don't knit well together…so start with a conversation about, where do you want to be? 

My position is always that you want to be on some kind of integrated, uniform platform, deployed on cloud, that constantly evolves, and which allows you to focus on business process change - because that's what's hard. 

My take

Supply chain is often an area of frustration with buyers. The tools that have been historically used have likely been pulled together and tweaked over the years in an attempt to react to changing needs and external market factors. We often hear that buyers are struggling to get the visibility and flexibility that they need, to move as quickly as they desire. In other words, supply chain management is quite reactive, rather than proactive. Chorley’s implied suggestion here is that Generative AI could help flip that script. The use cases highlighted here are interesting and we look forward to hearing from the customers that put them into practice. It’s early days for Generative AI, but we are starting to get a sense of how it may drive change in the coming months for the first adopters. 

A grey colored placeholder image