AI Agents: Powering the Next Leap in Supply Chain Modernization – A Roundup of Vendor Announcements

Originally Published May 2025, on ARCweb.com by Colin Masson

The modern supply chain is a crucible of volatility, dynamism, and distributed complexity. Navigating this landscape, which demands unprecedented multi-enterprise collaboration across intricate trading networks, requires a new breed of intelligent solutions. As I’ve explored in my research at ARC Advisory Group, most recently concerning the Industrial AI Revolution and how agentic AI will unleash the full industrial AI toolbox, we are on the cusp of a significant transformation. Building on earlier insights into unlocking supply chain potential with AI agents and multi-agent workflows, it's clear that AI Agents are poised to deliver on the full promise of AI in the industrial sector, particularly within our increasingly complex supply chains.

The recent flurry of activity and announcements at major industry events and from leading solution providers throughout early 2025 underscores this shift. AI Agents are no longer a futuristic concept but a rapidly materializing reality, offering tangible solutions to long-standing challenges, and pioneering industrial companies are already showcasing the potential.

The Foundational Pillars: Data Fabrics and Interoperability

For AI Agents to thrive and deliver transformative value, they require robust foundations. Two critical pillars stand out: Industrial-Grade Data Fabrics and seamless interoperability protocols.

My work on the core capabilities of an Industrial-Grade Data Fabric highlights how such a fabric is essential for powering AI infusion and modernization. It’s not just about data lakes or warehouses; it's about an active platform delivering services for connectivity, data ingestion, processing, contextualization, storage, modeling, and crucially, AI/ML model development and management. This fabric provides the operational engine for robust Industrial DataOps pipelines and supports the sophisticated AIOps solutions that modern supply chains demand.

Furthermore, as AI Agents become more prevalent, their ability to communicate and collaborate becomes paramount. This is where the rise of A2A (Agent2Agent) communication, completing the Industrial AI protocol stack alongside OPC UA and MCP (Model Context Protocol), is critical. MCP provides a standardized way for agents to connect to tools and data sources, acting like a universal connector. A2A then enables the vital peer-to-peer conversations, capability discovery, and task coordination between AI agents, irrespective of their origin or underlying framework. This three-layered stack (OPC UA for data access, MCP for agent-tool interface, and A2A for inter-agent collaboration) is fundamental for real-world Industrial AI use cases, from coordinating robotic fleets to optimizing intricate supply network responses.

AI Agents Take Center Stage: Insights and Industrial Showcases from Major 2025 Events and Announcements

The strategic importance of AI Agents, underpinned by these foundational elements, was a dominant theme in recent industry discussions and vendor announcements.

Kinaxis Kinexions 2025

At Kinexions 2025, Kinaxis emphasized AI-driven transformation, focusing on empowering "Supply Chain Heroes" to move "From Chaos to Control," as I covered in my ARC blog on the event. Key aspects of their AI agent strategy included:

• AI Agents within the Maestro Platform: Introduction and enhancement of AI agents integrated into their core Maestro platform, designed to assist supply chain professionals in navigating disruptions and automating critical tasks.

• Focus on Disruption Management and Automation: Agents are aimed at improving inventory management, risk mitigation, and overall planning efficiency by intelligently processing data and suggesting actions.

• Supply Chain Data Fabric with Databricks: A strategic partnership with Databricks was highlighted, underpinning the AI agent capabilities by forging a resilient Supply Chain Data Fabric for the Maestro platform, enabling more intelligent, data-driven decision-making.

• Agentic AI Framework for Customization: Kinaxis showcased its agentic AI framework, which allows users to create their own AI agents that can interact with operational data through natural language, empowering citizen developers.

• Real-world Impact (Merck & Co. Example): The potential of these agents was underscored by industrial applications, such as the mainstage discussion with Hanu Gadila of Merck & Co., who explored how AI agents could transform the handling of supply shortages, potentially cutting related manual effort by up to 80%. Industrial giants like ExxonMobil (Pioneer Award for S&OP transformation) and Schneider Electric (Impact Award for sustainability initiatives using Maestro) were also recognized for their innovative application of Kinaxis solutions, setting the stage for leveraging these new AI agent capabilities. Logistics Viewpoints (an ARC Advisory Group publication) also noted this trend.

Coupa Inspire 2025

Coupa’s Inspire 2025 event showcased a bold vision for AI Agents, particularly within spend management and multi-enterprise collaboration, aiming to create an "Autonomous Spend Management network." My ARC Advisory Group blog on Coupa Inspire 2025 detailed this strategy. Key AI agent announcements included:

• New Multi-Agent AI Portfolio: Coupa introduced a suite of AI agents designed to coordinate transactions and decisions between buyers and suppliers with increased autonomy.

• Navi™ Agents: These agents form a core part of the new portfolio, providing capabilities for advanced analytics, knowledge discovery, and other intelligent functions to assist users.

• "Bring Your Own AI Agent" (BYOAA) Framework: A significant move towards openness, this framework is designed to facilitate A2A collaboration by allowing customers and partners to develop and integrate their own specialized AI agents into the Coupa ecosystem.

• Leveraging a Massive Anonymized Dataset: Coupa emphasized that its AI agents would harness insights from its vast ($8 trillion) anonymized spend dataset, enabling powerful data-driven decision-making and benchmarking.

• Enhanced Category Management with Cirtuo Acquisition: The recent acquisition of Cirtuo, an AI-driven category management leader, is set to embed further intelligence into Coupa's Total Spend Management platform, likely enhancing agent capabilities in this area. Industrial leaders such as Schneider Electric, Lippert, GAF Materials, and Glencore were featured, highlighting advanced uses of Coupa's platform, which will benefit from these new agentic AI capabilities.

Blue Yonder ICON 2025

Blue Yonder ICON 2025 heavily emphasized its "Cognitive (R)evolution," launching new AI-infused Cognitive Solutions and, significantly, five critical AI agents:

• Inventory Ops Agent: To highlight supply/demand mismatches and recommend solutions.

• Shelf Ops Agent: To speed planogram editing via natural language.

• Logistics Ops Agent: To monitor transport, automate scheduling, and optimize routes.

• Warehouse Ops Agent: To dynamically allocate labor and optimize warehouse layouts.

• Network Ops Agent: To oversee multi-enterprise operations and manage disruptions. A Tariffs Agent is also reportedly in development. 

These capabilities are built upon the Blue Yonder Platform, leveraging the Snowflake AI Data Cloud, and are complemented by a new Supply Chain Knowledge Graph (developed with Snowflake and RelationalAI) to unlock insights from unstructured data. 

Blue Yonder also announced an Agent Activation Advisory service, in collaboration with Microsoft's AI Co-Innovation Labs and Azure AI Foundry, to expedite customer adoption. While specific industrial showcases of these newly launched agents are emerging, Blue Yonder stated these were "built in collaboration with customers." Companies like GXO Logistics (whose CTO discussed AI-driven supply chains on the mainstage), and ICONic Customer Award winners or presenters such as manufacturing firms Cummins, Hewlett Packard, International Paper, and Micron Technology, are prime examples of industrial players positioned to leverage these advanced agentic capabilities for further transformation. 

For additional insights into Blue Yonder's ICON 2025 and why supply chains must transform, see my colleague's detailed analysis here.

FourKites' Recent AI Advancements (Early May 2025)

In early May 2025, FourKites also publicly announced groundbreaking AI solutions designed to automate critical handoffs between transportation networks and warehouse operations, integrating with their Intelligent Control Tower™. Key innovations include:

• Yardworks AI: An evolution of their YardWorks suite, now infused with AI capabilities.

• Alan: An AI-powered "Digital Worker" dedicated to scheduling. Alan handles the entire appointment lifecycle across various communication channels (emails, portals), aiming to reduce manual scheduling workloads by as much as 50%.

• AutoGate AI: A computer vision solution that automates gate check-in and check-out processes, slashing check-in times from over ten minutes to under two minutes and enabling 24/7 secure operations by integrating with the FourKites platform for data validation. These new agents join FourKites' existing "Digital Workforce," which includes agents like "Tracy" for track-and-trace and "Sam" for supplier collaboration, signaling a comprehensive approach to automating supply chain tasks. Furthering its AI-driven inventory intelligence, FourKites also announced a strategic alliance with Chorus, an Alphabet X spinout. This partnership integrates Chorus's item-level sensor technology (Seeker™ labels) and AI tools with FourKites' Intelligent Control Tower™, aiming to create dynamic digital twins of inventory and enable autonomous inventory orchestration. 

Industrial companies like Arrow Electronics are actively leveraging FourKites' Intelligent Control Tower™ and Digital Workforce capabilities to achieve measurable outcomes, including fewer manual tasks and improved on-time delivery, showcasing the practical application of these advanced AI solutions.

SAP Sapphire 2025

SAP's Sapphire 2025 event marked a significant declaration of intent with its "Business AI" vision, as I detailed in my ARC blog, "SAP's Audacious AI Gambit: A Skeptic's Look at the Business AI Vision from Sapphire 2025." SAP is architecting an ecosystem where AI is woven into the fabric of the enterprise, focusing on delivering tangible AI value rapidly.

At the core is the "Business AI Flywheel," uniting Applications, Data, and AI.

• Data Revolution with Business Data Cloud (BDC): Resonating strongly with the Industrial-Grade Data Fabric concept, SAP's BDC, built on SAP Datasphere, aims to unify all enterprise data (SAP and non-SAP) within a single semantic layer. Strategic alliances with Databricks (enabling advanced analytics on context-rich SAP data) and Palantir Technologies (for seamless connectivity with Palantir Foundry and AIP) underscore a commitment to an open data ecosystem.

• Joule as the New Generative UI: SAP's AI copilot, Joule, is evolving beyond an embedded assistant to become an "omnipresent" generative UI for all enterprise interactions. The acquisition of WalkMe will make Joule follow users across SAP and third-party applications via the "SAP Action Bar for Joule". Integration with Perplexity AI will enhance Joule's ability to combine internal SAP data with public web knowledge for comprehensive answers.

• The Rise of AI Agents: SAP is heavily investing in AI agents designed to "reason and act," moving beyond simple automation to proactive problem-solving. Sapphire 2025 saw the announcement of numerous pre-built agents, including critical Supply Chain agents such as the Shop Floor Supervisor, Maintenance Planner, and Field Service Dispatcher, alongside agents for Finance, Spend Management, HCM, and CX, mostly slated for general availability in Q3/Q4 2025.

• Empowering the Ecosystem for Agent Development: Joule Studio (within SAP Build) will offer a Skill Builder and an AI Agent Builder for custom agent creation. The new AI Agent Hub in SAP LeanIX will provide a centralized governance layer for all enterprise AI agents. Furthermore, SAP's support for the Agent2Agent protocol with Google Cloud signals a future of interoperable AI ecosystems. 

This comprehensive strategy, underpinned by the SAP Business Technology Platform (BTP) and a strong push for customers to adopt a "Clean Core" via RISE with SAP and GROW with SAP, indicates SAP's determination to lead in enterprise AI.

Manhattan Associates Momentum 2025

Manhattan Associates used its Momentum 2025 conference to unveil a significant strategic shift, placing Agentic AI at the heart of its future roadmap. The company announced sweeping Agentic AI support within its Manhattan Active® solutions, positioning this as a generational leap. The core of this strategy involves embedding intelligent, autonomous digital agents capable of reshaping supply chain commerce execution.

Key announcements included:

• Five new out-of-the-box AI agents: Intelligent Store Manager: For natural language interaction and real-time optimization in store operations. Labor Optimizer Agent: To enhance workforce management through dynamic optimization and natural language. Wave Inventory Research Agent: Facilitating complex inventory research and optimization via natural language. Contextual Data Assistant: An evolution of Manhattan Active® Assist for surfacing relevant data and driving optimization. Virtual Configuration Consultant: To simplify application configuration and optimization using natural language.

• Manhattan Agent Foundry™: A groundbreaking platform engineered using Google Agentspace technology and the Vertex AI platform. This foundry is designed to empower customers and partners to build and deploy their own bespoke AI agents, fostering an open innovation ecosystem. The Foundry aims to support emerging industry standards like A2A and MCP protocols for interoperability. 

These capabilities, slated for general availability in Fall 2025, signal Manhattan's ambition to move beyond GenAI assistance to truly autonomous and adaptive supply chain operations, leveraging its cloud-native, microservice-based Manhattan Active® Platform. The expanded partnership with Google Cloud is also a critical enabler for this AI strategy.

The New Collaborative Paradigm: AI Agents in Volatile Supply Networks

The developments from these major vendors and the experiences of industrial end-users vividly illustrate how AI Agents are uniquely suited to address the core challenges of modern supply chains:

• Volatility and Disruption Management: As seen with Merck's exploration at Kinexions, Blue Yonder's purpose-built agents, FourKites' Alan for dynamic rescheduling, SAP's new suite of supply chain agents, and Manhattan's focus on real-time disruption resolution, these intelligent assistants can monitor vast data streams, predict potential disruptions, and proactively suggest or execute mitigation.

• Dynamic Decision-Making: The ability to rapidly process complex scenarios is paramount. The work by industrial leaders highlighted, coupled with the new agent capabilities for automated decisioning (e.g., FourKites' AutoGate AI, Blue Yonder's Ops Agents, SAP's reasoning agents, Manhattan's autonomous agents), underscores this AI-driven agility.

• Distributed Operations and Data Silos: Industrial Data Fabrics (Kinaxis with Databricks, Blue Yonder with Snowflake, SAP's Business Data Cloud) and unified data platforms (Coupa's platform, FourKites' Intelligent Control Tower, Manhattan Active® Platform) are crucial. AI Agents operating on these foundations can gain holistic visibility and orchestrate actions across the network.

• Multi-Enterprise Collaboration: This is where the true power of A2A communication (a principle SAP and Manhattan are supporting) and collaborative AI agent networks comes to the fore. Coupa's vision, Kinaxis's collaborative planning, Blue Yonder’s Network Ops Agent, FourKites' Digital Workforce, SAP's agent ecosystem, and Manhattan's Agent Foundry all point to more seamless inter-company collaboration.

• Complexity and Scale: Modern industrial supply networks involve immense variables. AI Agents, as explicitly offered by these leading vendors, can manage this complexity by distributing tasks, specializing, and coordinating efforts.

The Path Forward: Intelligent Automation and Orchestration in the Industrial Sector

The journey towards fully agentic supply chains is well underway, with tangible examples and robust roadmaps emerging from the industrial front lines and their technology partners. The emphasis is shifting from standalone AI tools to interconnected networks of intelligent agents capable of perception, reasoning, decision-making, and autonomous action. These agents, built upon robust data fabrics and communicating through standardized protocols like MCP and A2A (and increasingly, embedded within vendor platforms that support such concepts), will empower supply chain professionals in industrial enterprises, transforming their roles from manual task execution to strategic oversight and exception management.

The commitment from solution providers, coupled with the pioneering efforts of industrial leaders, signals that the pace of innovation is accelerating. For industrial organizations looking to build resilient, agile, and intelligent supply chains, embracing the potential of AI Agents is no longer a matter of if, but when and how. The groundwork laid by advancements in data infrastructure, interoperability standards, and the early successes of industry trailblazers is paving the way for this new era of supply chain modernization.

Engage with ARC Advisory Group

For ARC Advisory Group recommendations for navigating the AI Wars, closing the digital divide by embracing Industrial AI, assembling your Industrial-grade Data Fabric, and modern Industrial AI technology stack - and governing and guiding major decisions about enterprise, cloud, industrial edge, and AI software, please contact Colin Masson at cmasson@arcweb.com or set up a meeting with me, or my fellow Analysts at ARC Advisory Group.