How Procurement Leaders Can Use Agentic AI And AI Agents Together

Forbes

30-06-2025

Shaz Khan is the CEO and co-founder of Vroozi, a digital procure-to-pay platform.
Nearly every procurement leader I meet with voices the same challenge: doing more with fewer resources—without sacrificing compliance, quality or productivity. Organizations are increasingly turning to AI to help with this challenge, but many remain uncertain about how to embed these tools into their operations—specifically the distinction between AI agents and agentic AI. Understanding this difference is critical for procurement and supply chain leaders who want to stay competitive.
AI Agents: Process-Driven Task Handlers
AI agents require human interaction and specific instructions to accomplish a task or deliver an outcome. They operate within clearly defined parameters and output structured responses based on preprogrammed rules. We already interact with these agents daily—think of voice assistants like Siri or Google Assistant, which can provide weather updates, check traffic or make reservations.
In procurement, AI agents offer practical use cases. For example, a technician might scan a part number and type 'replacement motherboard'' into a chat interface. The AI agent could locate the item among preferred suppliers, provide pricing and delivery options and route the purchase order for approval—all triggered by a simple request.
AI agents can also automate invoice processing by scanning documents, transferring data, routing approvals and initiating payments—tasks that typically take 20 minutes per invoice. Their value lies in managing repetitive tasks without fatigue. They excel at transactional activities like creating purchase orders, assisting in supplier sourcing and onboarding, and conducting evaluations. However, AI agents have limited decision-making ability and require manual retraining by developers to improve over time.
Agentic AI: The Autonomous Orchestrator
While AI agents act as task handlers, agentic AI can offer more autonomy with the potential to operate entire business processes with minimal human oversight. Rather than requiring a person to make each decision, agentic AI could evaluate scenarios, consider alternatives and select the best course of action using real-time data.
One emerging use case is tail spend management. Most procurement leaders recognize that a small percentage of their suppliers generate the majority of their spending. Agentic AI could continuously monitor a high volume of suppliers, identify opportunities for consolidation, manage risk profiles and pinpoint alternative sources of supply—factoring in economic trends or policies and global disruptions. It could proactively recommend: 'These suppliers can offer your products at an average 6% lower cost, with faster delivery. Would you like to connect with them?"
How They Work Together
Agentic AI and AI agents do not have to be competing approaches—they're complementary. Organizations seeing the greatest impact might consider using both in tandem. For example, a chatbot (an AI agent) might receive the request: 'I need an Apple laptop for my new employee Sarah who is starting Thursday.' Agentic AI could then determine Sarah's needs based on her role and location.
Here, the AI agent manages the front-end conversation, while the agentic system could handle the decision-making and process execution in the background.
Procurement leaders who want to successfully leverage these technologies should begin with the following steps:
1. Educate On AI Terminology And Frameworks: Partner with subject matter experts and invest in training that covers agentic AI and generative AI.
2. Commit To Learning Through Experimentation: Form innovation teams to tackle real issues and iterate quickly. Accept that not every pilot will succeed.
3. Invest In Data Quality: AI needs clean, structured data to function properly. Poor data leads to inaccurate pricing and missed delivery targets.
4. Focus On Governance And Security: Autonomous processes increase the need for robust data security and compliance oversight.
5. Start With Clear Processes: AI performs best when integrated into structured workflows and well-defined procurement procedures.
Empowering Procurement Professionals
This evolution isn't about replacing procurement professionals—it's about enhancing their capabilities. The companies leading this shift aren't simply rebranding legacy tools with AI labels. Instead, they're building intelligent systems that handle massive data flows and generate actionable insights guide real decision-making.
AI value lies in relevance and role-specific recommendations. A specialty chemicals buyer, for instance, might receive alerts about global volatility and sourcing recommendations tailored to their segment. Over time, the confusing terminology—agentic AI, agent AI, generative AI—will fade. What matters is how these tools are applied to reduce costs and solve real-world problems, freeing up human experts to focus on strategic, high-impact work.
By starting with clear problem definitions, improving data quality and thoughtfully combining both agentic and agent-based approaches, procurement teams can move beyond the pressure of "doing more with less" and start doing more with strategic precision and purpose.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?