The post Supreme Court Strikes Down Trump Emergency Tariffs appeared first on Logistics Viewpoints.

A year ago, when people talked about AI in supply chain, they mostly meant chatbots that could answer questions about shipment status or generative models summarizing reports. Useful stuff, but incremental. That’s changed fast.

What’s emerged over the past twelve months is a different class of AI altogether. AI agents can now execute multi-step workflows autonomously, coordinating across systems, making decisions based on real-time data, and acting on those decisions without waiting for a human to click “approve.” They read shipping documents, cross-reference contracted rates, flag discrepancies, and initiate dispute processes. They monitor inbound shipments, detect delays, adjust dock schedules, and notify downstream teams. They do this continuously, across thousands of transactions per week.

I don’t want to belabor the point here. If you’ve been paying attention to reports like the Bain Technology Report or McKinsey’s State of AI survey, you already know the trajectory. The technology is real. The harder question for logistics and supply chain leaders is what it means for how their organizations operate.

The Opportunity: Collapsing Operational Silos

Here’s the argument I want to make plainly: An agentic AI operating layer, built on supply chain data, will collapse the organizational silos that have defined how large shippers run their businesses for decades.

The technology isn’t magic. Supply chain data happens to be the connective tissue between departments that have historically operated as if they had nothing to do with each other.

Finance needs delivery confirmation to trigger early payment discounts. Procurement needs carrier performance data to update scorecards. Customer service needs real-time order status to respond to penalty claims. Production planning needs inbound ETAs to adjust manufacturing schedules. Insurance needs shipment documentation to process claims.

All of these decisions happen in different departments, in different systems, managed by different teams. But they all start with real-time data about shipments, orders, inventory, and deliveries.

For years, the handoffs between “supply chain knows something” and “another department acts on it” have been manual. Someone pulls a report. Someone else verifies it. A third person takes action in a different system. That’s how most companies still operate. And most of the time, it’s a reaction to a disruption rather than proactive alignment across functions.

An AI operating layer changes that equation. When agents can ingest supply chain data in real time, apply business rules, and execute actions across enterprise systems, those manual handoffs disappear. A delayed inbound shipment doesn’t wait for someone to notice it in a report and then email the warehouse. The agent detects the delay, recalculates the dock schedule, and notifies the facility team before anyone opens a spreadsheet.

Supply Chain Data as a Trigger

At FourKites, we’ve deployed AI agents that handle specific operational functions autonomously. One monitors shipments around the clock, investigates delays, and coordinates with carriers. At Coca-Cola, it cut response times for “where’s my truck” queries from 90 minutes to seconds. Another handles supplier collaboration, reading shipping documents and creating tracking records automatically. A third manages customer and vendor scheduling, reducing team workload by half at facilities like US Cold Storage.

But the more interesting development is what happens when you extend beyond traditional logistics workflows. Things like automatically validating freight invoices against contracted rates and actual service levels. Or accelerating payment cycles by identifying early discount opportunities tied to delivery confirmation.

More than “visibility” use cases, these automations extend to finance, procurement, warehouse operations, and customer service. But they all depend on supply chain data as the trigger. This is increasingly how leading shippers are thinking about their technology stack — connecting supply chain platforms directly to ERPs, CRMs, and financial systems so that operational data can trigger action in those systems without manual intervention. Gartner’s 2025 Supply Chain Top 25 highlighted this move toward autonomous, cross-system orchestration as one of the defining characteristics of the highest-performing supply chains globally.

The workflow executes in another function, but the intelligence that drives it originates in the supply chain. That’s what makes supply chain the starting point for an enterprise-wide AI operating layer, not the boundary of it. So the question becomes what it takes to actually stand up an operating layer like this.

What’s Required to Build It

Let me be honest about what it takes, because I think there’s been too much hand-waving in the market about AI transformation.

Start with the data foundation. An operating layer is only as good as the data flowing through it. For shippers, that means having a real-time view of what’s happening across your supply chain network, not a batch-updated dashboard that’s six hours stale. You need live shipment status, carrier performance history, order-level tracking, facility throughput data, and the system integrations to connect it all. If your data is fragmented across disconnected point solutions, the AI has nothing meaningful to work with.

Focus on proven workflows, don’t automate broken ones. This is the hardest part, and it’s where most companies stall. McKinsey’s 2025 State of AI survey found that 88% of organizations now use AI in at least one business function, but only about 6% are capturing meaningful enterprise-wide value from it. The biggest differentiator between those groups is workflow design. For example, a freight invoice audit that currently involves three people touching a spreadsheet could be replaced by an agent that cross-references the contracted rate, validates the service level against tracking data, and flags only genuine discrepancies for human review.

Build for orchestration across systems, not within one system. Here’s where the general-purpose AI platforms fall short. Many of them are good at connecting to your systems and building automations for whatever you throw at them. But they don’t have context from an external network that reveals impacts to your operations. They start with your data alone.

A supply chain operating layer starts with your data plus the operational intelligence from a broader network: which carriers perform well on which lanes, how delays in one region tend to ripple to facilities in another, and what distinguishes a genuine exception from normal variability. That context is what allows agents to act, not just surface alerts.

The Pace of Change

I also want to acknowledge something that too many people are glossing over. This stuff has moved unbelievably fast. The industry has been talking about AI agents for over a year now, but they’ve only become truly viable in production settings in the past few months. The underlying model capabilities, the integration tooling, the orchestration frameworks. All of it has matured at a pace that’s genuinely difficult for any organization to keep up with.

Jason Lemkin at SaaStr recently described what’s happening in enterprise software as a structural budget reallocation. IT spending is growing modestly overall, but AI budgets are absorbing a disproportionate share. Application counts are flat. Seat-based growth is under pressure. Companies aren’t spending more on software. They’re spending differently, and they’re spending on outcomes.

For supply chain automation specifically, you don’t need a multi-year transformation program to get started. The modular architectures that exist today make it possible to deploy production-grade agents in weeks rather than quarters. And platforms like FourKites’ Loft now make it possible to build and configure AI agents around your specific business rules, SOPs, and system integrations — not a one-size-fits-all workflow.

But to get the most ROI, you must first understand the workflows that consume the most manual effort and document the SOPs that govern how your teams handle exceptions, validate data, and communicate across functions. That’s the raw material that AI agents need to operate effectively.

The technology is ready. Whether your organization has done the foundational work to take advantage of it is a different question, and it’s the one worth spending time on.

By Matt Elenjickal, CEO, FourKites

The post The AI-Powered Operating Layer Has Arrived, and Your Supply Chain Is Where It Starts appeared first on Logistics Viewpoints.

The conversation around AI in supply chain is evolving.

We have moved beyond proofs of concept and isolated copilots. The central question is no longer whether an agent can summarize a planning report or respond to a transportation exception.

The real question is this:

Can AI systems operate across domains, under governance, and at production scale?

That is not a model question. It is an architectural one.

A layered approach built around Agent to Agent communication, A2A, and the Model Context Protocol, MCP, provides a structured way forward. Not as features. As infrastructure.

Coordination vs Capability: The Foundational Separation

At a high level, the pattern is straightforward:

A2A provides the coordination layer

MCP provides the capability layer

This separation is more consequential than it first appears.

Without it, agent systems collapse into distributed monoliths characterized by:

Embedded business logic inside agents

Hardcoded integrations

Tight coupling between workflows

Limited extensibility

With proper separation:

Orchestration remains distinct

Execution logic is encapsulated

Capabilities are modular

The system can evolve without structural rewrites

This is the difference between experimentation and operational architecture.

A2A: The Coordination Layer

A2A allows agents to discover and communicate with one another through standardized interfaces. Each agent publishes an Agent Card describing:

Capabilities

Acceptable request types

Invocation parameters

Other agents can discover and invoke these capabilities without tight coupling.

For supply chain leaders, the implications are concrete:

A Transportation Agent calls a Compliance Agent

A Supplier Risk Agent coordinates with a Financial Exposure Agent

An Order Promising Agent interacts with a Warehouse Capacity Agent

The objective is not simply inter agent messaging. It is controlled interoperability across domains without embedding vendor specific logic inside every workflow.

This is how specialization scales.

MCP: The Capability Layer

If A2A governs how agents talk, MCP governs how they act.

The Model Context Protocol standardizes how tools, structured data, and predefined prompts are exposed to agents. Rather than embedding all operational logic inside the agent itself, MCP allows capabilities to be modular and discoverable.

In a supply chain context, MCP tools might include:

get_atp_snapshot

quote_spot_rate

screen_restricted_party

check_wave_capacity

generate_trade_documents

Adding a new compliance requirement or operational rule does not require rewriting orchestration logic.

It requires deploying a new tool.

This distinction enables:

Extensibility instead of fragility

Controlled evolution of capability

Separation between business intent and operational mechanics

The Layered Architectural Pattern

This model resolves into three defined roles:

Orchestrator Agent

Translates high level business intent into sequenced tasks

Maintains visibility into the overall objective

Specialist Agents

Execute domain specific responsibilities

Encapsulate transportation, compliance, sourcing, fulfillment, or risk logic

MCP Tool Layer

Provides granular, reusable operational capabilities

Exposes APIs, data services, and rule checks in modular form

The separation is deliberate:

Orchestrators own intent and sequencing

Specialists own execution logic

Tools remain modular and reusable

This ensures:

Business intent remains readable

Execution remains encapsulated

Capabilities remain composable

A Practical Scenario

Consider a high value customer order at risk of service failure.

Business objective: Recover service without eroding margin.

The orchestrator agent decomposes the goal into:

Assess constraints and risk

Generate recovery options

Validate feasibility

Execute and monitor

Through A2A, it coordinates:

Order Promising Agent

Transportation Agent

Compliance Agent

Warehouse Agent

Customer Communication Agent

Each specialist invokes MCP tools relevant to its domain, such as:

Allocation rules

Spot rate quotes

Compliance screening

Capacity checks

CRM case creation

Now introduce change:

A new emissions reporting requirement

A new supplier expedite option

In a layered architecture, these changes require:

Registering a new tool

Or introducing a new specialist agent

They do not require redesigning orchestration logic.

That is structural resilience.

Architectural Advantages

A layered A2A and MCP model enables:

Dynamic Discovery

New agents can join the ecosystem

Orchestration logic does not require rewrites

Composable Capabilities

Specialists assemble behavior from modular tools

Logic is not embedded permanently inside agents

Separation of Intent and Execution

Business goals remain governable

Execution details are isolated and replaceable

Adaptability

New requirements are met through composition

Structural reengineering is minimized

For enterprises operating globally, these are prerequisites, not enhancements.

Governance Is Not Optional

As agents discover tools and access systems, governance becomes central.

Enterprise grade deployment requires:

Strong identity and authorization controls

Tool level access management

Full decision logging and auditability

Human approval gates where required

Deterministic fallback behavior

Autonomy without control increases operational and regulatory risk.

Layered architecture enables governance. It does not replace it.

Coexistence with Deterministic Workflow Engines

This model does not eliminate traditional workflow orchestration platforms.

Those systems remain essential for:

Reliability

Scheduling

Observability

SLA enforcement

The layered model complements them:

Workflow engines provide deterministic backbone and operational control

A2A enables flexible coordination across agents

MCP standardizes capability exposure

The result is adaptability without sacrificing operational discipline.

The Bottom Line

Supply chain AI will not be determined by who deploys the most capable standalone model.

It will be determined by who builds systems that:

Coordinate effectively across domains

Incorporate new capabilities without architectural rewrites

Maintain control under regulatory pressure

Avoid recreating monoliths in distributed form

A2A and MCP represent a structured attempt to provide that foundation.

The post Architecting Agentic Operations for Supply Chain – A Practical View of A2A and MCP appeared first on Logistics Viewpoints.