Status Quo
Global supply chains can’t flow on software from the last century.
Modern supply chains are more global, complex, and dynamic than ever. Yet the infrastructure powering them is still largely built on technology from another era:
Email (1971). OCR (1974). EDI (1979). ERPs (1990). PDF (1993).
Over the past decades, supply chain tools have improved, but the underlying model has barely changed. Systems remain siloed. Communication happens via emails, PDFs, and Excel sheets. Connecting a new partner takes months. Data is fragmented and inconsistent. Humans act as the glue between systems — entering data, chasing confirmations, correcting errors, and handling exceptions.
The result is constant friction. The worst part is that this has been accepted as the norm, because companies have operated within these constraints for decades.
The Future
Supply chains will become autonomous.
AI is a different technology shift from the ones that came before it. Personal computers, the internet, and mobile all created new platforms that humans could build on top of. AI, on the other hand, creates entirely new economic actors: intelligent agents that can buy, sell, plan, reconcile, coordinate, and make decisions on behalf of companies.
The impact on global trade seems inevitable. Supply chains will become autonomous. Workflows will move from human-operated to agent-executed, and within a few years every company in the supply chain will have fleets of agents executing tasks on their behalf.
The Gap
Intelligence is here. The infrastructure is not.
LLMs provided intelligence and agents brought autonomy. But today’s supply chain stack was built for humans, not agents. Data is messy, access is fragmented, context is implicit, and guardrails are unclear.
Agents can work around these constraints for individual use cases, but they cannot reliably run entire supply chains on top of them. To reach true autonomy, the underlying infrastructure itself needs to become AI-native.
Having agents operate without the right infrastructure will actually create more friction, not less. This is especially dangerous in supply chains, where one error can cost millions and where complexity, workarounds, and customisations are the norm.
The opportunity is therefore not just to automate the existing operating model, but to redesign it for an agent-native world.
Our Convictions
A new operating model for supply chain automation
1. Software becomes invisible
Today, software is still human-operated: it has interfaces, needs to be configured, and relies on users to keep processes moving. We believe that model is ending. Software will become self-executing and increasingly invisible — humans define intent, constraints, and exceptions, while software continuously acts in the background.
2. Automation becomes democratised
As models improve, the cost of building workflow automations goes to zero. Every company becomes a technology company and every team member becomes a builder. The people closest to a workflow will be the ones automating it — tailored to their needs, with full control over their data and know-how.
3. Context becomes shared
Today, data is trapped in siloed systems of record — hard to access, inconsistent, constantly reconciled between companies. Every system and agent will instead operate on accurate, shared, real-time context. Instead of exchanging different versions of reality, everyone operates on the same shared truth.
Procuros
AI-native supply chain connectivity
The future operating model for supply chains requires an entirely new connectivity infrastructure: one that makes companies, systems, and data accessible to agents by default. At Procuros, we call this AI-native supply chain connectivity. It moves beyond the stack built around EDI, OCR, PDFs, and emails to create a new foundation for autonomous supply chains.
Procuros provides instant connectivity with any trade partner and trusted data that is validated, harmonised, and easily accessible. It acts as a shared system of truth between partners.
It is an infrastructure built for an agent-native world, giving agents the data and system access they need to act autonomously. It also enables users to automate virtually any workflow in their supply chain and eliminate the friction they have lived through for decades.
Everything flows
We’re building for a great future
Our bet for the future is that supply chains will become programmable. Software will run itself, every user will automate processes, and agents will operate on a shared context across company boundaries.
Soon agents will handle the busywork, while humans spend time on valuable tasks. We believe the underlying infrastructure will be the missing piece to make supply chains truly autonomous.
We are not building software for the world as it is. We are building it for the world as it should be. A world where everything flows.