How AI Is Re‑Architecting Industrial Procurement and Supply Chain

This article is sponsored by Arkestro and was written, edited, and published in alignment with our Emerj sponsored content guidelines. Learn more about our thought leadership and content creation services on our Emerj Media Services page.

Enterprise procurement leaders are operating in environments characterized by increasing supplier complexity, data intensity, and external volatility. As organizations scale, procurement functions are expected to support cost control, supply continuity, and informed decision‑making under uncertainty.

The GAO has documented the performance gap between strategic and reactive buying in concrete terms. Leading companies strategically manage about 90% of their procurements and report annual savings of 10% or more, while the federal agencies GAO reviewed were managing about 5% of their spend through strategic sourcing efforts.

The operational pressure on procurement functions is intensifying. The Hackett Group’s 2025 Key Issues Study — which benchmarks procurement operations across 97% of the Dow Jones Industrials and 89% of the Fortune 100 — found that procurement workloads are projected to increase 10% while budgets grow just 1%, creating a 9% efficiency gap. According to the same study, 64% of procurement leaders expect AI and generative AI to fundamentally transform their roles within five years.

The OECD’s 2025 Government at a Glance report clearly frames the institutional imperative: modernizing procurement systems is now considered essential, with a central emphasis on digital technologies to increase transparency, responsiveness, and data-driven decision-making.

The case for transformation is real, but the results are not automatic: procurement value depends heavily on data quality, process discipline, and organizational readiness.

Emerj recently spoke with senior leaders across life sciences, energy, and heavy industry to understand how procurement is shifting from reactive buying to a science‑driven, AI‑enabled strategic function. Featured voices include Rob DeSantis, CEO and Co‑founder of Arkestro; Madhav Madaboosi, Head of Digital Transformation in Future Midstream and Strategy at bp; Mike Shin, Chief Supply Chain Officer at Trinity Rail Industries; Damion Nero, Head of Data for U.S. Medical at Takeda Pharmaceuticals; and Shreyas Becker, Head of AI & Data Products at Sanofi.

These conversations surfaced five procurement‑specific insights that illustrate how AI is reshaping sourcing, supplier management, and operational decision‑making.

Procurement’s risk posture as the hidden adoption barrier: The same caution that protects supply continuity and supplier relationships also slows AI transformation.

Dynamic evaluation of sourcing options in unstable markets: Clearer insight into viable local and regional suppliers helps procurement maintain continuity when global routes become unreliable

Continuous, risk‑based monitoring at scale: Replacing static surveys and episodic assessments with continuous, exception‑based monitoring preserves visibility as supplier networks grow and allows leaders to focus on material risk signals rather than overwhelming volumes of data.

Bottom‑up adoption as the key to procurement transformation: Demonstrating quick, frontline value through simple, targeted proofs of concept builds the credibility needed to secure long‑term investment.

Frictionless sourcing as the foundation for predictive procurement: Removing manual data gathering and quote analysis accelerates cycle times and lets teams focus on higher‑value decisions.

Procurement’s risk posture as the hidden adoption barrier

Guest: Rob DeSantis, CEO and Co-founder, Arkestro

Episode: Enabling Strategic Procurement with AI, From Frustration to Foresight – with Rob DeSantis of Arkestro

Expertise: Procurement, Supply Chain Transformation, E-procurement, SaaS Leadership, Strategic Sourcing, Enterprise Software Scaling

Brief Recognition: Rob DeSantis is the CEO and Co-founder of Arkestro, bringing over 30 years of experience to the intersection of enterprise software and supply chain operations. He previously co-founded Ariba, a foundational leader in e-procurement, and served as a member of the early executive team at LinkedIn. Throughout his career, DeSantis has specialized in leveraging emerging technologies to drive measurable step-function value and earnings-per-share impact for global corporations.

Rob DeSantis begins the conversation by drawing attention to a dynamic that rarely gets named explicitly inside large enterprises: procurement’s instinct to protect continuity often slows the adoption of new technology, even when the upside is clear.

Finance pushes for earnings impact; procurement protects supply stability. Those incentives diverge the moment AI enters the conversation.

DeSantis explains that procurement’s caution is not cultural hesitation; it is a structural requirement of the role. A failed experiment can jeopardize supply availability, damage supplier relationships, or expose the organization to compliance risks. That reality makes new technology feel less like an opportunity and more like a potential disruption.

He captures the tension directly:

“Supply chain resists what finance is willing to bet on: every transformative wave — from the internet to cloud to AI — hits a ‘turbo‑lag’ of fear and conservatism, yet the organizations that overcome it capture not incremental gains, but step‑change value that ultimately makes the risk unavoidable.”

– Rob DeSantis, CEO and Co-founder at Arkestro

From there, DeSantis outlines the forces that reinforce this posture:

Continuity exposure: Any disruption threatens production schedules and customer commitments.

Supplier relationship sensitivity: Procurement avoids moves that could destabilize long‑standing partnerships.

Operational overload: Teams are already stretched by data volume and legacy tools, leaving little bandwidth for experimentation.

Finance, by contrast, sees AI through the lens of step‑function savings rather than operational risk. As DeSantis notes, finance leaders are often more willing to take the leap because the potential impact is so large.

This misalignment creates what he calls “turbo lag” — the multi‑year delay between a new technology’s arrival and procurement’s willingness to adopt it. He’s watched the same pattern unfold with the internet, the cloud, and now AI.

Rob emphasizes that overcoming this lag requires acknowledging procurement’s risk posture rather than working around it. Without that recognition, even high‑ROI opportunities stall before they begin.

His strategic takeaway is clear: AI adoption in procurement will not accelerate until leaders actively manage, not ignore, the function’s inherent risk aversion.

Dynamic evaluation of sourcing options in unstable markets

Episode: Scaling Drug Manufacturing from Clinical Trials to Commercial Production – with Shreyas Becker of Sanofi

Guest: Shreyas Becker, Head of AI & Data Products: Manufacturing & Supply, Sanofi

Expertise: Manufacturing AI, Supply Chain Resilience, Tech Transfer, Reasoning Models, Life Sciences Operations, Data Product Management

Brief Recognition: Shreyas Becker is the Head of AI and Data Products for Manufacturing and Supply at Sanofi, leading the integration of reasoning-based AI into high-stakes pharmaceutical manufacturing. He specializes in accelerating the tech transfer phase and building robust data foundational layers to improve supply chain resilience and manufacturing throughput.

What stands out in Becker’s perspective is how quickly he dismisses the idea that today’s volatility is temporary. In his view, the last several years didn’t break the system; they revealed what was already fragile. And once that becomes clear, the question shifts from How do we optimize the old model? to Why are we still using it?

He points out that supply chain teams have spent more than a decade tuning processes for stability that no longer exists. Tariffs, geopolitical shifts, and pandemic‑era disruptions forced organizations to confront the limits of global dependency.

Instead of squeezing out another percentage point of efficiency, teams suddenly had permission, and necessity, to rethink where and how they source.

He puts it plainly:

“For the last 15 years, we’ve been talking about optimization. We rarely talk about redesign. Now we are talking about it. Shocks give you an opportunity to redesign an entire thing so you could leapfrog a lot of the small challenges and make significant gains.”

– Shreyas Becker, Head of AI & Data Products: Manufacturing & Supply, Sanofi

That redesign inevitably changes the sourcing map. Some materials still require specialized global setups, but many others don’t. Becker notes that for commoditized ingredients, suppliers can now differentiate on quality and reliability, not just cost; a shift that makes regional and local options far more competitive than they were five years ago.

AI becomes the mechanism that makes this rethink possible. Not because it automates the old process, but because it can evaluate conditions that the old process was never built to handle. Earlier systems struggled with edge cases; newer models can reason through unfamiliar scenarios, weigh trade-offs, and surface alternatives that weren’t previously visible.

The result is a sourcing function that behaves differently:

It doesn’t wait for a disruption to reconsider suppliers.

It doesn’t assume global routes are the default.

It doesn’t treat volatility as an exception.

Becker’s point is that AI doesn’t just help procurement react faster — it helps procurement see entirely different options, especially when the environment is unstable. Volatility becomes a design input, not a crisis.

Continuous, risk‑based monitoring at scale

Episode: Leveraging Data to Scale Drug Development Globally – with Damion Nero of Takeda

Guest: Damion Nero, Head of Data for U.S. Medical, Takeda Pharmaceuticals

Expertise: Precision Medicine, Data Science, Deglobalization Strategy, Clinical Development, Real-World Evidence, Pharmaceutical Analytics

Brief Recognition: Damion Nero is the Head of Data for U.S. Medical at Takeda Pharmaceuticals, where he leads the strategic application of machine learning and real-world data to global drug development. With over 15 years of experience in precision medicine, he specializes in navigating the transition toward regionalized supply chains and localized data sourcing to maintain profitability in a fragmented global market.

Damion Nero describes a supply‑chain environment where the ground moves faster than the systems built to track it.

Tariffs appear before anyone knows how they’ll be collected, ports stall without warning, and policy shifts outpace the infrastructure meant to enforce them. In that kind of landscape, the traditional rhythm of supplier oversight — scheduled reviews, periodic surveys, episodic assessments — simply can’t keep up.

The deeper issue, in Nero’s view, is that the global model for which those tools were designed is dissolving. The long era of U.S.‑backed free trade is giving way to a more fragmented, regionalized system. New blocs are forming, old alliances are weakening, and supply routes that were stable for decades are becoming unreliable. Pharmaceutical companies can no longer assume that a global supplier will remain accessible, compliant, or even operational.

That shift forces a different posture:

• Supply lines have to shorten.

• Redundancies have to be built locally.

• Procurement teams need visibility that doesn’t arrive weeks or months after conditions have changed.

Continuous monitoring becomes less about sophistication and more about survival — a way to detect the early signs of disruption before they cascade into shortages, delays, or market loss.

Nero makes the stakes clear, and the quote lands best when it closes the section:

“What comes after is really sort

[truncated for AI cost control]