The Most Important Business Rules in Your Supply Chain May Be Sitting in a PDF

The $100 million agreement nobody's systems understand

Imagine two multinational companies entering into a $100 million supply agreement.

Legal negotiates the contract. Procurement agrees on pricing and quantities. Logistics establishes delivery requirements. Finance negotiates payment terms. Executives approve the arrangement.

After weeks or months of negotiations, the agreement is signed. Then the work of making it operational begins.

The legal contract is filed away, perhaps in a contract management system, a SharePoint repository, someone's email, or even a physical filing cabinet.

Meanwhile, IT begins the work of making the agreement operational.

ERP systems must be configured. EDI trading partner profiles must be created. Pricing conditions, payment terms, tolerances, shipping instructions, and invoicing requirements must be entered into multiple systems.

In effect, an agreement that two companies have already negotiated must now be interpreted and implemented again, this time by software specialists.

We have digitized the transactions, but frequently left the agreements governing them disconnected from execution.

And the consequences are measurable.

The research tells a troubling story

A 2021 study by EY and the Harvard Law School Center on the Legal Profession uncovered significant weaknesses in how organizations manage contracts.

Among participating organizations:

• 90% reported difficulty locating contracts because they lacked appropriate technology or processes.
• 78% did not systematically track contractual obligations.
• 71% said they lacked technology to monitor contracts for deviations from standard terms.
• 49% lacked a defined process for storing contracts after execution.

These are findings from a 2021 study, not measurements of every business today. Nevertheless, they illustrate the longstanding disconnect between negotiating commercial terms and operationalizing them.

Source: EY, The General Counsel Imperative: How Does Contracting Complexity Hide Clear Profitability? (2021)

The financial implications are significant.

World Commerce & Contracting’s August 2025 contract-management research reports average contract value erosion of 8.6%.

Value erosion represents the difference between the commercial value an organization expects from its contracts and the value ultimately realized.

This is not a measure of recoverable ERP errors, and it would be misleading to suggest that automation could recover the entire 8.6%.

But it demonstrates how much economic value can be at stake when contractual commitments aren't effectively translated into business outcomes.

Source: WorldCC, Contract Management Whitepaper (August 2025)

Why onboarding takes weeks or months

Businesses have spent decades making electronic commerce possible.

EDI standards such as ANSI X12 and EDIFACT allow companies to exchange purchase orders, acknowledgments, shipment notices, invoices, and payment information.

But establishing the connection is only part of the challenge.

An enterprise integration provider estimates that traditional EDI onboarding can take six to twelve weeks per trading partner, largely because of manual mapping, configuration, and repeated testing.

That is a vendor-reported estimate, not an independently established industry average. Actual timelines vary considerably by complexity and existing infrastructure.

Source: Cleo, How to Onboard EDI Trading Partners Faster

The deeper challenge is operational alignment.

Both companies must determine how their systems will interpret and execute their trading relationship.

Which prices apply? What quantities are permissible? Which delivery dates are acceptable? What happens if a shipment is late? Can an order be changed? Who is authorized to approve an exception?

Many of these answers already exist in the trading agreement.

Yet the associated rules frequently must be recreated in ERP configurations, EDI implementation guides, spreadsheets, and operating procedures.

Every new partner, contract amendment, and pricing change can introduce another opportunity for the systems and the agreement to diverge.

What if the contract became machine-readable?

Imagine a fundamentally different approach. A company uploads an existing supply agreement into an intelligent platform.

AI analyzes the document and extracts the operationally relevant provisions:

• Trading parties, products, prices, currencies, and discounts.
• Order quantities, tolerances, delivery windows, and fulfillment conditions.
• Payment terms, early-payment discounts, and credit arrangements.
• Acceptance criteria, service levels, penalties, and remedies.
• Amendment procedures, cancellation conditions, and approval requirements.
• Authorized entities, individuals, and software agents.

The extracted information is structured into a machine-readable Executable Agreement, or EA.

AI extraction is not the same as legal interpretation or authorization. Ambiguous terms, conflicting amendments, and business rules requiring judgment must be resolved by authorized people before becoming enforceable system controls.

The original legal contract remains authoritative. The EA becomes its approved operational representation, retaining traceability to the underlying contractual language.

The agreement can now participate directly in the transaction lifecycle.

From detecting exceptions to preventing them

Consider a manufacturer that has agreed to purchase a component for $100 per unit, allowing a 2% price tolerance.

A purchase order arrives electronically with a price of $108.

The EDI message is technically valid.

The ERP may accept it.

But the price violates the commercial agreement.

In a traditional environment, that discrepancy may only emerge after the supplier has confirmed the order, shipped the goods, and issued an invoice.

The buyer's accounts payable department discovers the mismatch during reconciliation.

Procurement gets involved. Sales operations investigates. The supplier disputes the invoice adjustment. Someone searches for the latest contract amendment.

An issue that could have been identified at order creation has now affected multiple departments across two enterprises.

With an approved Executable Agreement and a governance engine, the discrepancy can be identified as soon as the transaction becomes visible.

The governance engine compares the price against the agreed tolerance.

It determines that the order exceeds the permitted variance.

Depending on configured authority and system integration, it can flag, hold, reject, or escalate the transaction for resolution.

The parties address the difference before it creates further financial and operational consequences.

Instead of discovering an exception downstream, we can identify the conditions that create it upstream.

That is a fundamental shift in how companies conduct business.

What is the economic opportunity?

The benefits extend beyond the direct cost of processing invoices.

APQC's cross-industry benchmarking reports a median $6 cost to process an accounts payable invoice, based on data from 5,846 organizations. That figure covers the broader accounts payable process, including personnel, systems, and overhead. It is not the cost of an exception alone.

Source: APQC, What Is the Average Cost to Process an Invoice? (October 2026)

But invoice-processing costs are only one component of the opportunity.

A contract mismatch can trigger manual investigation, credit notes, rejected invoices, delayed payments, inventory adjustments, and disputes spanning both enterprises.

To illustrate the potential economics, consider a hypothetical company with:

Under these assumptions, the company could avoid handling 30,000 exceptions annually.

At $30 per exception, that represents $900,000 in potential annual gross handling-cost savings.

These values are scenario assumptions, not published industry benchmarks or demonstrated NeurWare performance.

Additional savings might arise from reducing trading-partner configuration effort, eliminating contractual pricing errors, accelerating dispute resolution, and improving cash flow.

For example, if 100 new or changed trading partner configurations require 80 labor hours each at $100 per hour, reducing that workload by 35% would yield another $280,000 of annual capacity value.

Together, these two hypothetical improvements represent $1.18 million in annual gross efficiency opportunity, before implementation costs, operating expenses, and any overlap between savings categories.

The appropriate way to establish actual ROI is to measure baseline exception rates, handling costs, onboarding effort, preventable discrepancies, and resolution times using a customer's own operational data.

The important opportunity isn't a universal savings percentage.

It's the possibility of removing unnecessary work from millions of transactions by understanding the rules governing those transactions before they become problems.

Digitizing the space between companies

For decades, enterprise technology has focused on optimizing the internal operations of individual companies.

ERP platforms organize the business.

EDI and integration networks exchange information between businesses.

Contract lifecycle management platforms manage the creation, negotiation, and storage of agreements.

All are essential.

Yet a significant opportunity remains in connecting these capabilities across organizational boundaries.

Imagine an intelligent operating layer that understands the trading relationships between companies.

One that observes existing transactions, reconstructs their underlying business processes, compares them against approved commercial agreements, and identifies operational and financial opportunities.

One that provides a shared, permissioned foundation for validating transactions, resolving exceptions, coordinating changes, and eventually supporting autonomous commerce.

This is the vision behind NeurWare: Make Intercompany Intelligent.

See it.

Understand the intercompany transactions already flowing through existing EDI networks, APIs, and enterprise systems.

Use process intelligence and a Control Tower to reveal how business is actually conducted, including recurring exceptions, operational bottlenecks, and financial opportunities.

Govern it.

Connect those transactions to the commercial agreements that define the relationship.

Validate prices, quantities, delivery conditions, payment obligations, trading identities, and delegated authority against approved rules.

Identify exceptions at the earliest practical point, and enforce decisions where the participating organizations have granted the necessary authority.

Orchestrate it.

Coordinate actions across organizational boundaries.

Help buyers and suppliers resolve exceptions, manage amendments, route approvals, and synchronize business outcomes.

Use intelligent agents where appropriate, supported by deterministic rules, human oversight, and an auditable record of decisions.

These capabilities can coexist with existing enterprise applications and integration infrastructure rather than requiring companies to replace them.

The agents are coming. Are our agreements ready?

There is another reason this matters.

AI agents are increasingly capable of performing activities previously reserved for human employees.

Soon, more agents will be involved in purchasing, order management, supplier communications, contract administration, and financial workflows.

But an AI agent that can create a purchase order is not necessarily authorized to create one.

An agent that can negotiate pricing must still operate within the authority granted by its organization.

And two agents representing different companies must understand the rules governing their interaction.

Who are they acting for?

What are they permitted to do?

Which agreement applies?

What are the financial limits?

When must a human approve the transaction?

Without reliable answers, autonomous commerce risks accelerating the same inconsistencies that already generate exceptions and disputes.

Machine-readable Executable Agreements, combined with identity verification, delegated authority, deterministic governance, and auditable transaction evidence, could help establish the infrastructure that makes cross-enterprise agentic commerce trustworthy.

The next frontier of digital transformation

We have invested enormous resources in digitizing what happens inside enterprises.

We have also built sophisticated networks for exchanging information between them.

Yet the agreements that define commercial relationships often remain disconnected from the digital systems responsible for executing them.

The research demonstrates that managing those agreements remains difficult, contractual obligations are not always systematically monitored, and significant commercial value can be lost through ineffective contract management.

AI now gives us new ways to extract and structure contractual knowledge.

Governance technology gives us ways to validate transactions against that knowledge.

And intelligent orchestration creates opportunities to resolve problems before they become disputes.

The next breakthrough in intercompany commerce may not be another transaction format, another integration, or another ERP module.

It may be giving our existing systems a shared understanding of what the companies actually agreed to.

Because the contract shouldn't stop working once it's signed.

It should be where intelligent commerce begins.