Bill Foy: How AWS is Modernising Supply Chains With AI
According to Bill Foy, Director and APJ Automotive Sales Leader at AWS, 94% of companies have been affected by supply chain disruptions – with legacy systems causing real-time visibility gaps.
“Whether it's a port strike, an earthquake or a natural disaster, most companies don’t even know a problem exists until it’s downstream,” he says.
Managing this disconnected data can be particularly challenging for original equipment manufacturers (OEMs), Bill explains.
“OEMs struggle with data scattered across Excel spreadsheets, mainframes – all sorts of different systems,” he says. “They don’t communicate together, so there is no unified model.”
To overcome this, Bill says that companies need to take a more unified and predictive approach to managing their supply chains.
“OEMs that are really moving forward are moving from a reactive, batch-driven operation to real-time AI-powered, customer-centric supply chains,” he says.
Legacy systems slow AI adoption
“For almost all of the OEMs I work with, their data is trapped in legacy systems,” Bill says. “They’re running mission-critical supply chain operations in mainframes that are 30 to 45 years old.”
Much of this software is written in COBOL – a software language very few people still operate in.
“There’s no documentation,” says Bill. “The reverse engineering alone can take hundreds of hours, just to understand what’s in those systems.”
Having these legacy systems in place can also make it harder for companies to adopt new systems.
Bill shares that, while 87% of manufactures see AI as an important tool, only 8% have researched the mature stage of implementation.
“One of the biggest challenges with data being trapped in legacy systems is that they weren’t designed for real-time data exchange,” he says. “This disconnected, siloed data remains one of the biggest barriers for AI adoption.”
Modernising Toyota’s manufacturing operations
However, leaving behind legacy systems and scaling AI out of an initial pilot phase can be done successfully – as seen in AWS’s partnership with Toyota Motor North America.
“Toyota North America’s legacy mainframe systems managed 90% of their supply chain operations,” Bill says. “And a mainframe outage would basically halt car sales entirely.”
To evolve this system, AWS deployed AWS Transform for Mainframe, using AI to analyse millions of lines of code. This, Bill says, was the first Agentic AI service designed to modernise mainframe workloads at scale.
“After validation from Toyota’s own COBOL engineers, the teams produce complete, high-quality documentation in a single day. This is work that would have taken them months to do.”
On the business side, this transformation helped shift Toyota from legacy build-to-stock push models to a modernised, customer-centric ‘pull’ model.
“This really allowed the seamless integration between sales and product manufacturing, fundamentally enhancing the customer experience and unlocking new profit opportunities,” says Bill.
Building future-ready systems
“We’re seeing a shift from AI as a tool to AI as an autonomous operating layer for manufacturing and supply chains,” Bill says.
He suggests that future systems will autonomously orchestrate demand sensing, inventory positioning and logistics optimisation within defined guardrails – resolving supplier descriptions “before a human even sees the alert”.
However, he advises that manufacturing leaders first ensure they have a “unified data fabric” before they apply AI models.
“Manufacturing organisations’ first priority should be connecting the manufacturing floor data to the ERP system, to supplier networks, to logistics platforms,” Bill says. “AI is only as good as the data it can access in real time.”
