Uploaded July 2026 | Updated September 2026, 2 weeks ago
Procurement may be digitised on paper.
But the real supplier decision is often won or lost before ERP sees it.
I’m joined by Spencer Penn, CEO and co-founder of LightSource, a direct procurement AI platform working on one of the messiest decision areas in manufacturing: how engineering, suppliers, finance, and procurement actually align when products, prices, specifications, and risks keep moving.
This matters now because manufacturers are under pressure from cost volatility, supplier fragility, regulatory demands, labour constraints, and shorter product cycles. Speed is no longer a nice advantage. It is becoming the difference between resilience and exposure. If sourcing decisions are still scattered across Excel, email, and institutional memory, visibility arrives late — and late visibility is often just expensive hindsight.
What changed my thinking was Spencer’s point that procurement AI should not be framed primarily as headcount reduction. In direct procurement, the value at stake is too large. Individual procurement managers may be influencing decisions that create or destroy hundreds of millions in value. The better question is: how do you augment those decisions before cost creep, supplier risk, or launch delays become embedded?
Spencer also shares a sharp example from an automotive EV programme where part of the vehicle was sourced through LightSource and part was not. The result: 25% shorter sourcing cycles and 37% less cost creep. That is not a dashboard story. That is a business outcome.
There is also a useful distinction between systems of record and systems of action. ERP stores the truth. AI agents may increasingly help move the work — supplier discovery, RFX preparation, quote follow-up, market-change signals, and post-award risk.
This is for supply chain, procurement, manufacturing, operations, risk, and sustainability leaders who need better decisions before disruption, cost, or carbon exposure hardens into business impact.
If you’re dealing with this on the ground, I’d like to hear how you’re handling it.
Listen to more episodes: resilientsupplychainpodcast.com
Subscribe for more supply chain resilience, procurement, sustainability, risk, and operations analysis.
Chapters / Timestamps
00:00 – When ERP sees procurement too late
00:32 – Why digitised sourcing still hides manual risk
01:57 – LightSource and the direct procurement AI problem
02:27 – The gap between PLM, ERP, and suppliers
05:23 – Why direct materials are still under-tooled
08:42 – Speed becomes the new manufacturing advantage
12:36 – AI assistants versus agents that act
18:34 – Why procurement should be augmented, not replaced
23:19 – Three no-regrets AI moves for CPOs
26:40 – Where AI should attack sourcing pain first
29:18 – What changes when teams leave Excel and email
33:26 – Faster sourcing and lower cost creep
Procurement may be digitised on paper.
But the real supplier decision is often won or lost before ERP sees it.
I’m joined by Spencer Penn, CEO and co-founder of LightSource, a direct procurement AI platform working on one of the messiest decision areas in manufacturing: how engineering, suppliers, finance, and procurement actually align when products, prices, specifications, and risks keep moving.
This matters now because manufacturers are under pressure from cost volatility, supplier fragility, regulatory demands, labour constraints, and shorter product cycles. Speed is no longer a nice advantage. It is becoming the difference between resilience and exposure. If sourcing decisions are still scattered across Excel, email, and institutional memory, visibility arrives late — and late visibility is often just expensive hindsight.
What changed my thinking was Spencer’s point that procurement AI should not be framed primarily as headcount reduction. In direct procurement, the value at stake is too large. Individual procurement managers may be influencing decisions that create or destroy hundreds of millions in value. The better question is: how do you augment those decisions before cost creep, supplier risk, or launch delays become embedded?
Spencer also shares a sharp example from an automotive EV programme where part of the vehicle was sourced through LightSource and part was not. The result: 25% shorter sourcing cycles and 37% less cost creep. That is not a dashboard story. That is a business outcome.
There is also a useful distinction between systems of record and systems of action. ERP stores the truth. AI agents may increasingly help move the work — supplier discovery, RFX preparation, quote follow-up, market-change signals, and post-award risk.
This is for supply chain, procurement, manufacturing, operations, risk, and sustainability leaders who need better decisions before disruption, cost, or carbon exposure hardens into business impact.
If you’re dealing with this on the ground, I’d like to hear how you’re handling it.
Listen to more episodes: resilientsupplychainpodcast.com
Subscribe for more supply chain resilience, procurement, sustainability, risk, and operations analysis.
Chapters / Timestamps
00:00 – When ERP sees procurement too late
00:32 – Why digitised sourcing still hides manual risk
01:57 – LightSource and the direct procurement AI problem
02:27 – The gap between PLM, ERP, and suppliers
05:23 – Why direct materials are still under-tooled
08:42 – Speed becomes the new manufacturing advantage
12:36 – AI assistants versus agents that act
18:34 – Why procurement should be augmented, not replaced
23:19 – Three no-regrets AI moves for CPOs
26:40 – Where AI should attack sourcing pain first
29:18 – What changes when teams leave Excel and email
33:26 – Faster sourcing and lower cost creep
