For many procurement professionals, the daily reality of managing multi-billion-dollar supply chains still looks like a digital version of operations in the 1990s. Despite the rapid evolution of global trade, a staggering 75% of procurement teams still rely on a combination of fragmented spreadsheets and endless email chains to manage direct materials.

Spencer Penn, Co-Founder and CEO of LightSource, knows this alarming status quo better than most. Before founding the AI-powered strategic sourcing platform, Spencer was a leader at Tesla, where he helped launch the Model 3 programme. It was there, amidst the pressure of scaling from 1,000 vehicles per week to millions, that he witnessed the breaking point of traditional procurement.

"One of the big challenges that we faced was that we were sourcing 30 billion of direct materials on Excel spreadsheets and emails," Spencer recalls. "I sourced millions of parts for the Model 3 with email chains that contained hundreds of messages. As we neared production, I knew there had to be a better way."

The missing system of record

In the modern manufacturing landscape, the distinction between direct and indirect procurement is often where the digital transformation journey stutters. While indirect spend, the "tail spend" of office supplies and travel, has seen significant innovation, direct materials remain trapped in manual processes.

Spencer describes this disparity through a vivid analogy: "The way I describe it is that indirect procurement is like UPS when it picks up a package. It has to route it and drop it off. It’s a one-time flow and it’s done. And then direct materials are much more like the plumbing in a building. You have to set it up right."

For companies like Tesla, or LightSource’s current clients such as Yum! Brands and Bombardier, the "plumbing" of direct materials is the lifeblood of the business. Yet, while sales teams have Salesforce and engineers have CAD, procurement has lacked a unified system of record. "LightSource is akin to how Salesforce works for a sales team," Spencer explains. "Salesforce helps sales teams find customers and sell to them and then procurement is the same thing in reverse. We’re finding vendors and we’re buying from them. So instead of a sales motion, we’re running a sourcing motion."

The shift from generative to agentic AI

As we move into 2026, the conversation around artificial intelligence in the supply chain has shifted. The industry is moving past the "AI washing" phase – where legacy vendors simply layer chatbots over decades-old infrastructure – and into the era of Agentic AI.

Spencer is quick to distinguish between the two. While Generative AI (Gen AI) is reactive, producing text or summaries based on a prompt, Agentic AI is proactive. "I think the AI space has had a second step change... particularly everything related to agents and agentic workflows," he notes. "Gen AI gives you an output from a prompt, but with agentic workflows you input a prompt and get a set of actions or a sequence of actions."

For a procurement team, this isn't just about writing a better RFP. Rather, it’s about having what Spencer calls "small armies of AI interns" that can execute multi-step processes. These agents can handle BOM (Bill of Materials) management, track thousands of line items as engineering revisions occur and perform real-time tariff and cost calculations.

However, Spencer warns against the "capability overhang”, where technology outpaces the actual utility of the products being sold. "Our responsibility at LightSource is making AI useful and application-oriented. If you tell someone to use ChatGPT to make their job easier, you’re asking procurement people to suddenly be creative about technology use, which is a totally new skillset that is very different from the actual job itself."

Navigating the ‘Year of Supply Chain Redesign’

The timing for this technological leap is critical. Spencer refers to the current era as the "year of supply chain redesign”. While the post-pandemic years were focused on stabilisation and normalisation, today’s challenges are more structural.

"There is another subset of customers that have just started popping up for us, which are customers that are worried about what’s going to happen with Trump tariffs," Spencer observes. "And so they’re thinking, now that we’re going to have an entirely new trade regime, now we have to go back and redesign our entire supply chain."

This redesign isn't just about finding the lowest price, a tactic that Spencer argues is no longer sufficient. Modern procurement involves managing complex trade-offs between cost, reliability, sustainability and regional risk. Whether it is nearshoring, onshoring, or dual sourcing, the complexity of these decisions requires a level of data intelligence that a spreadsheet simply cannot provide.

"It’s not enough anymore to just put quotes out to bid and choose the lowest-price supplier," says Spencer. "Today’s advanced procurement teams need to consider complex trade-offs in a dynamic environment that’s always changing."

The 5% lever: margin vs. growth

For C-level executives, the argument for modernising procurement often boils down to a simple mathematical reality regarding margins. Spencer points out that for many businesses, improving procurement is a more efficient path to profitability than aggressive sales growth.

"There are two big levers for improving your margin: you could double sales and double your profit, or you could reduce your costs through better sourcing and better negotiation by like 5% and then you’ve achieved the same outcome. I think that’s a much easier path in my view."

Despite this, the transition to digital tools often faces internal resistance. Spencer identifies executive buy-in as both the greatest driver and the most significant obstacle to progress. "When there’s a leader that is very comfortable with what they already know, they’re living in their comfort zone and they don’t really want to reinvent the wheel... then it’s always rolling a rock up a hill."

Augmentation, not replacement

A common fear in the procurement community is that AI will eventually replace the human buyer. Spencer’s philosophy, however, is firmly rooted in augmentation. He believes the goal of technology should be to "make the easy things easier and the hard things possible."

"There have been two thoughts about AI’s future in procurement: replace people or enable people to do more. At LightSource, we focus on augmenting people," he says. "The answer isn’t to replace procurement, but augment and help them work faster and better."

This augmentation is particularly vital during the New Product Introduction (NPI) phase. With 80% of cost variability determined by the time a design is frozen, the ability for procurement to collaborate with engineering in real-time is the difference between a successful launch and a financial struggle.

Looking ahead: spec-to-scale

As LightSource prepares to unveil its "spec-to-scale" multi-agent systems, Spencer’s vision remains clear: procurement must move from being seen as a "necessary evil" to a "critical business driver."

By moving away from "procurement as usual", the cycle of manual entry, IP leakage risks and fragmented data, teams can finally focus on the strategic work they were hired for.

"In the short term, technology adoption is a popularity contest, but in the long term, it’s based on actual performance," Spencer concludes.

For a procurement audience facing a "totally new trade regime" and the physical era of AI, the performance of their digital tools may soon be the most interesting and important topic on the boardroom agenda.