How agentic AI fits into Supply Chain Management – Diginomica

Agentic AI and Sustainable Development Goals in Supply Chain Management

Introduction

Agentic AI has emerged as a significant advancement in Supply Chain Management (SCM) this year, driven by the capabilities of Large Language Models (LLMs) to enhance automation. Many SCM platform vendors are integrating LLM-enhanced agentic features, while Decision Intelligence providers improve supply chain processes through advanced data management and AI techniques. These developments align with several Sustainable Development Goals (SDGs), including SDG 9 (Industry, Innovation, and Infrastructure) and SDG 12 (Responsible Consumption and Production), by promoting innovation and efficiency in supply chains.

Challenges in Current SCM Workflows

Despite progress, significant gaps remain in SCM workflows due to fragmented data from multiple systems and inconsistent data timing. This fragmentation undermines trust among stakeholders, who often rely on experience rather than data-driven insights, potentially misaligning with broader organizational goals. Addressing these issues supports SDG 8 (Decent Work and Economic Growth) by improving decision-making and operational efficiency.

“The platforms also took a considerable amount of time to implement and deliver value, with some nearing twenty-four months for companies to start seeing the payback. Many enterprises weren’t able to do complete data cleansing in this time, falling short of their goals. Finally, many of the platforms have limited collaboration with humans within the organization and with suppliers. This can lead systems to mis-model supply chains with a lack of context or a lack of validation with suppliers.”

Agentic AI as a Tactical Solution

Agentic AI offers faster, cost-effective SCM solutions without requiring platform changes. It enables employees and suppliers to collaborate on models and optimize supply chains, fostering partnerships that contribute to SDG 17 (Partnerships for the Goals).

New Agentic Approaches in SCM

Limitations of Legacy Automation

While Robotic Process Automation (RPA) bots and process intelligence tools have advanced, they often fail in unknown situations, causing supply chain delays. Agentic AI, powered by LLMs, adapts dynamically to diverse scenarios, reducing maintenance costs and complexity.

“RPA bots can work well when everything is known, and there aren’t issues or opportunities to optimize. However, RPA bots often break and stop when they encounter unknown situations, and this can slow your supply chain.”

Agentic AI Capabilities

• Handling mismatched purchase orders by identifying alternative suppliers and placing orders within budget and time constraints.
• Alerting downstream processes and notifying customers proactively.
• Reducing the need to pre-identify all failure types, simplifying solutions and controlling costs.

“Unlike RPA bots, you don’t need to identify all the failure types and paths up front, simplifying your solutions and keeping your costs in check.”

Emerging Use Cases and SDG Alignment

Proactive Disruption Management

Agentic LLMs analyze unstructured data such as emails, weather reports, and labor disputes to predict shipping delays 3-7 days in advance, enabling timely production replanning and alternative delivery arrangements. This supports SDG 9 and SDG 12 by enhancing supply chain resilience and sustainability.

Contract Analysis and Compliance

LLMs review contracts to identify non-standard or non-compliant terms, improving regulatory adherence and supporting SDG 16 (Peace, Justice, and Strong Institutions).

Product Improvement through Customer Feedback

Analysis of unstructured customer reviews helps companies adjust product features and perceptions, contributing to SDG 12 by promoting responsible production and consumption.

Research and Innovation

Ongoing research focuses on fully automating decision processes, preventing AI hallucinations, and uncovering causal relationships in supply chains. These innovations advance SDG 9 by fostering industry innovation.

“Implementors need to take care in balancing where they allow the agent to identify this type of solution and allowing the agent to make a decision based on novel ideas. If you find the middle ground, you will find agents that can come up with solutions that can be approved via human in the loop processes that might not have been surfaced via traditional or prior methods.”

Establishing New Guardrails for Trustworthy AI

Decision Authority and Human-in-the-Loop

Defining clear decision authority levels ensures that AI agents operate within acceptable risk parameters, preventing cascading errors. This approach aligns with SDG 16 by promoting accountable and transparent institutions.

Data Provenance and Traceability

Tracking data origins and decision-making processes enables organizations to audit and improve AI-driven decisions, supporting SDG 9 and SDG 16.

“Poor decisions will happen and, just like from bugs in traditional software, being able to trace why that decision happened is important to help improve your solution.”

Feedback Loops and Stability Controls

Implementing feedback mechanisms allows AI agents to learn from interactions and reinforce correct behaviors, enhancing system stability and continuous improvement.

Recommendations for Implementation

Start Small and Measurable

1. Select simple, measurable problems that consume employee time but add limited value.
2. Consider multi-agent workflows to address interconnected processes.
3. Assess data quality and cleanliness to prepare internal systems for agentic AI integration.

Develop Guardrails and Infrastructure

• Implement retrieval-augmented generation (RAG) solutions with vector databases for compliance.
• Ensure citations and policy engines enforce business and regulatory rules.
• Maintain schema consistency for agent inputs and outputs.
• Store data to support continuous improvement and auditing.

“If you need to ensure compliance, you’ll probably want a retrieval-augmented generation (RAG) solution… You’ll also want to make sure that inputs to and outputs from your agent match your schema.”

Human-in-the-Loop and Feedback Integration

Define roles for human oversight in agentic workflows, capture agent recommendations and decisions, and track outcomes to refine AI performance continuously.

Addressing Data Governance and Readiness

Data governance remains a critical challenge. Many organizations overestimate data readiness due to platform use, while underlying issues such as duplicates and contradictions persist. Auditing data governance practices and ensuring data lineage are essential steps to support SDG 16 and SDG 9.

“Many companies think their data is ready because they use platform X and their data is cleaned appropriately and well understood. In some cases, internal policies have failed or been ignored, and data in downstream systems becomes harder to match to the system of record.”

Future Outlook and Industry Impact

LLM-powered agentic AI has the potential to revolutionize supply chains within the next year, enabling analysts to identify disruptive opportunities beyond traditional models. Industry initiatives may develop SCM-specific LLMs, skill systems, and standard operating procedures embedded in agent infrastructure, fostering innovation aligned with SDG 9 and SDG 17.

“Agentic AI is going to fill needs that the largest players in SCM started to fulfill, but fell short of. The large SCM solutions aren’t going away, but the complexity of working in them will shrink as we see more and more agents built and deployed.”

Conclusion

Agentic AI presents a genuine opportunity to enhance supply chain efficiency and sustainability by connecting existing tools and platforms more effectively. A measured approach emphasizing decision auditing, data governance, and feedback loops can ensure alignment with Sustainable Development Goals, particularly SDG 8, 9, 12, 16, and 17, fostering responsible, innovative, and collaborative supply chain ecosystems.

1. Sustainable Development Goals (SDGs) Addressed or Connected

1. SDG 9: Industry, Innovation and Infrastructure
   • The article discusses innovations in Supply Chain Management (SCM) using agentic AI and Large Language Models (LLMs) to improve industrial processes and infrastructure efficiency.
2. SDG 8: Decent Work and Economic Growth
   • Improving supply chain workflows and decision-making supports economic growth and productivity, while also considering the impact on employees and encouraging value addition.
3. SDG 12: Responsible Consumption and Production
   • Enhanced supply chain transparency, data governance, and compliance with regulations contribute to responsible production and consumption patterns.
4. SDG 17: Partnerships for the Goals
   • The article highlights collaboration between employees, suppliers, and AI agents, as well as industry initiatives and vendor partnerships to innovate SCM solutions.

2. Specific Targets Under Those SDGs

1. SDG 9 Targets
   • 9.4: Upgrade infrastructure and retrofit industries to make them sustainable, with increased resource-use efficiency and greater adoption of clean and environmentally sound technologies.
   • 9.5: Enhance scientific research, upgrade technological capabilities of industrial sectors, particularly in developing countries.
2. SDG 8 Targets
   • 8.2: Achieve higher levels of economic productivity through diversification, technological upgrading and innovation.
   • 8.8: Protect labour rights and promote safe and secure working environments for all workers.
3. SDG 12 Targets
   • 12.6: Encourage companies, especially large and transnational companies, to adopt sustainable practices and to integrate sustainability information into their reporting cycle.
   • 12.4: Achieve environmentally sound management of chemicals and all wastes throughout their life cycle.
4. SDG 17 Targets
   • 17.16: Enhance the Global Partnership for Sustainable Development, complemented by multi-stakeholder partnerships that mobilize and share knowledge, expertise, technology and financial resources.

3. Indicators Mentioned or Implied to Measure Progress

1. Data Quality and Cleanliness Indicators
   • Measures of data fragmentation, duplication, contradictions, and data provenance tracking to assess data readiness and governance.
2. Implementation and Payback Timeframes
   • Time taken for SCM platforms or agentic AI solutions to deliver measurable value (e.g., 24 months for payback).
3. Decision Accuracy and Confidence
   • Tracking recommendations, decisions, and outcomes made by AI agents to evaluate system accuracy and reliability.
4. Human-in-the-Loop Engagement
   • Indicators related to the level of human oversight in decision-making processes to ensure trust and safety.
5. Supply Chain Disruption Prediction
   • Ability to identify potential disruptions 3 to 7 days in advance using LLMs analyzing unstructured data.
6. Compliance and Regulatory Adherence
   • Use of retrieval-augmented generation (RAG) solutions with citation and policy engines to ensure business and regulatory compliance.

4. Table of SDGs, Targets, and Indicators

Source: diginomica.com