GFT Tech: Why Digital Supply Chains Need Trust
Manufacturers spend millions on supply chain software. Yet, teams still run operations on manual spreadsheets and phone calls.
Fragmented systems create conflicting data. When executive dashboards tell the wrong story, factory workers stop trusting the tools.
Brandon Speweik is Head of Manufacturing at GFT Technologies US, working with leading manufacturers across the country to help improve operations with technology.
Ignasi Barri leads AI and data globally at GFT and aims to help move projects beyond the pilot phase.
Brandon and Ignasi share their expertise on digital supply chains.
See the full story in the June 2026 edition of Manufacturing Digital.
What is a digital supply chain in practical terms?
Brandon: A digital supply chain connects traditional supply chain systems into a single, holistic platform. This means that when something changes, such as a shipment delay, a component failure or a supplier shutdown, the right people find out fast enough to act on it, rather than after the damage is done.
Most manufacturers already have software managing different parts of their supply chain. The problem is that each system tells a different story. The production schedule says a job is ready to go, but the factory floor knows it isn't because a part is on hold. The supplier has a third version of events, and the people making decisions are working from whichever version landed in their inbox last.
For example, we worked with a manufacturer whose planning system showed a production run as fully resourced and on schedule. The line stopped because a critical component had failed inspection days earlier, but that information sat in a separate quality system that didn't feed into the scheduling software. By the time the production team found out, they'd already missed a delivery window.
A digital supply chain connects those systems so everyone is working from the same picture. That shift enables manufacturers to move from finding out after the damage is done to seeing problems as they develop, and it's where the value lies. It's harder to achieve than most companies expect, because it's not a technology problem, it's a data problem.
Why do so many digital supply chain initiatives stall?
Brandon: Most digital supply chain initiatives fail because companies skip the hardest part. They want an AI layer that creates executive dashboards and control-tower capabilities that are easy to demo in a boardroom. But they don't want to tackle the foundational work of ensuring all their underlying systems feed consistent, accurate data into a single shared view.
Manufacturers typically run separate systems for production scheduling, logistics, supplier management and quality tracking, and those systems rarely speak the same language. Most manufacturers have plenty of data, but what they don't have is truth.
The problem is that if a manufacturer's production scheduling system and factory floor are telling different stories, all that dashboards do is amplify messy data. The dashboard looks impressive until someone acts on it and the information turns out to be wrong. After that happens once or twice, people stop trusting it and go back to spreadsheets and phone calls.
The other issue is ownership. A supply chain transformation touches procurement, production, logistics, quality and IT, and in most organisations, nobody owns all of those together. So you end up with a lot of individual projects that never add up to a real capability. This often leads to pilot fatigue, where strong early results never scale because the organisational model was never built to sustain them.
How do you establish a solid digital foundation across systems?
Brandon: The starting point is getting systems to agree with each other rather than replacing them. Manufacturers have invested heavily in the software that manages production, tracks inventory and monitors quality, but ensuring those systems feed consistent, trusted information into a shared view is where the real work lies.
What we work toward with clients is what I'd call operational truth: a single picture that shows what was planned, what is actually happening, and whether the work met the required standard. Those three things sound simple, but they're often disconnected. The planning system shows intent. The factory floor shows reality. Quality data shows the outcome. Connecting all of them is the foundational work.
Once the data foundation is solid, AI becomes genuinely useful. But before, AI is essentially producing confident-sounding answers based on contradictory information. The companies that get the most out of AI in their supply chains are the ones that did the data work first.
Where is AI delivering value in supply chains right now?
Ignasi: Right now, AI is helping manufacturers prioritise and focus their energy on what matters most. A manufacturing supply chain generates an enormous amount of data, and employees spend a significant part of their day manually sorting through alerts and signals to determine what actually needs their attention. AI is very good at cutting through that noise to surface the issues that need a decision and filtering out the ones that don't.
We're also seeing strong results in document-heavy processes. Manufacturers deal with a constant flow of supplier communications, quality records, engineering change notices, compliance documents and more. AI can extract relevant information from those documents and insert it into the workflow where it's needed, rather than leaving it buried in an inbox or a shared drive.
The pattern in every successful deployment we've seen is the same: humans are still in the loop. It's making the people running it faster and better informed. The moment you manufacturers try to automate judgment calls without the data foundation to back it up, they’re in trouble.
How do you expect AI to deliver value in supply chains in a year's time?
Ignasi: Right now, most AI in supply chains is helping people understand what's happening. The next step is helping them act on it.
Today, when something goes wrong, such as a supplier missing a delivery or a quality issue shutting down a line, a team of people has to manually work through the impact, pull together the options, get the right approvals and coordinate a response. That process can take hours or days. AI is getting to the point where it can significantly compress that cycle by flagging the problem, mapping out the downstream impact, surfacing realistic alternatives and getting the right people moving faster.
This isn't the same as a fully automated supply chain. In most manufacturing environments, especially regulated ones, a person still needs to make the final call. But the coordination work of moving between systems, chasing information and running scenarios by hand is where AI will take on a much bigger role. The companies best positioned for this are the ones already doing the data work today.
What are the biggest roadblocks to adopting digital supply chains?
Brandon: Fragmented data is the biggest roadblock. Most manufacturers have a patchwork of systems that were never designed to work together, and getting a consistent picture across all of them is difficult. It’s an organisational challenge because a supply chain transformation touches almost every department. When no one is accountable for the whole thing, the result is a series of disconnected projects that each solve a pain point but never add up to a real capability.
Additionally, the roadblock that surprises clients the most is trust. If the factory floor has learned that the planning system is wrong more often than not, people stop using it as their source of truth. They develop workarounds like informal check-ins, their own spreadsheets and tribal knowledge. Getting the data right and convincing people to trust it again are two separate problems, and the second one is often harder than the first.
What innovations are coming next, and what needs to happen before they work?
Ignasi: The most interesting near-term development is AI that doesn't just surface information but helps coordinate action. Instead of flagging a disruption, it maps the downstream impact, pulls up response options and helps move the right people toward a decision faster. That shift from insight to coordination is where a lot of the next wave of value will come from.
Further out, there's real potential in scenario planning, which means using supply chain data to simulate how a disruption would play out before it fully develops, so teams can evaluate their options before they're in crisis mode. That could genuinely change how manufacturers approach resilience planning.
But both of those things depend on a data foundation that actually reflects what's happening in operations. We talk to a lot of manufacturers who want to jump straight to the advanced capabilities, digital twins, autonomous orchestration, predictive everything, before the foundation is in place. The conversation almost always comes back to the same question: Do your systems currently give you a reliable picture of what's happening? If not, the advanced capabilities won't deliver what you're hoping for. The most exciting innovations aren't a shortcut around the foundational work. They're built on top of it.
See the full story in the June 2026 edition of Manufacturing Digital.
