Tracing Procurement’s AI Evolution from Chatbots to Agentic AI

Procurement operations often navigate complexities, from external market volatility to intricate internal systems within SAP environments. These challenges can hinder process engagement and lead to non-compliant spending. However, Artificial Intelligence (AI), is increasingly simplifying these processes and elevating procurement’s strategic value.

During a recent SAP Ariba webinar, Barri Horn, Director of Product Marketing for AI for SAP Ariba and SAP Fieldglass’ Strategic Procurement Portfolios, noted that SAP aims to embed AI across its end-to-end portfolio to “make every process more efficient and every decision more data-driven.”

The Road to Co-Pilots

Procurement’s road to AI began with basic chatbots, which, while having limited functionality, paved the way for Generative AI (GenAI). Citing PwC data, Pierre Mitchell, Chief Research Officer and Managing Director of Spend Matters, highlighted GenAI’s transformative potential during the webinar. He explained that over two-thirds of companies believed that GenAI would significantly change how they create and deliver values. Horn concurred, stating, “It’s not hype. It’s a real game changer.”

GenAI’s strength in the SAP context is its ability to train on existing business data in solutions like SAP Cloud ERP, SAP Cloud ERP Private, SAP Ariba or SAP Fieldglass. This enables assistants to understand business environments, support multiple languages, and manage diverse queries without extensive IT scripting.

The next stage featured co-pilots that are designed to augment human users. SAP’s AI copilot, Joule, exemplifies this by allowing natural language interaction with business processes. Users can inquire about savings targets or request guidance on disruptions, and Joule will understand their intentions and initiate tasks like approvals.

From Assistant to Agent

However, agentic AI represents a new frontier for procurement processes, with intelligent agents managing complex, multi-step workflows independently. According to a Spend Matters whitepaper, these agents achieve user-defined objectives through planning, reasoning, and tool use, with human oversight. Mitchell explained that AI agents can act by understanding goals and context. “They can reason, and, over time, they can learn,” he said.

According to Horn, agentic AI raises the bar for procurement functions. It can adapt, reason, and execute tasks autonomously and manage processes with a goal. For example, an SAP finance agent, a supply chain agent, and a procurement agent can collaborate on complex scenarios.

This AI-first strategy aims to embed intelligence into workflows. Mitchell noted, “The biggest cost in procurement is the opportunity cost of spending your time creating POs and chasing transactions versus actually being able to do the strategic work that you need to do.” With agents, procurement teams are free of these tasks and can focus on more strategic ones.

Bridging the Gap

Still, integrating AI with existing systems is a critical technical aspect. Mitchell pointed out that to close this gap, organizations must tie the world of language models to the world of data models. “That convergence is really around knowledge and knowledge models,” he said. Techniques like Retrieval Augmented Generation (RAG) and knowledge graphs are key to grounding AI responses in actual business data, minimizing errors, and making AI output reliable within the SAP landscape.

And SAP is actively developing these areas. Horn said: “We’ve been rolling out Joule to support autonomous sourcing and price negotiations, among other things.”

Finally, the traditional operating model for ERP systems is evolving from scattered, often hidden data that was a huge hurdle for procurement teams. AI, coupled with solutions like SAP Business Technology Platform (BTP), is helping organizations overcome this barrier by enabling complete, reliable data for impactful decision-making.

What This Means for SAPinsiders

For procurement organizations, evaluating AI tools like Joule involves assessing SAP integration, data transparency, and the roadmap for agentic capabilities. For SAPinsiders, this means:

• Leverage AI for strategic focus: Recognize that AI assistants and agents are evolving to handle tactical tasks, freeing up procurement professionals for higher-value strategic activities. As Mitchell highlighted, AI can help mitigate the opportunity cost of spending time on non-strategic work.
• Understand data’s central role in leveraging AI: The effectiveness of AI in SAP environments hinges on connecting AI models to structured business data. Focus on solutions that employ robust methods like RAG to ensure AI outputs are accurate and contextually relevant.

Engage with emerging SAP AI tools: Explore and experiment with new SAP AI capabilities like Joule and agent development platforms. SAP’s advancements in these tools were highlighted during the recently concluded SAP Sapphire 2025 event in Orlando, where SAP and its users demonstrated how these tools are becoming more intuitive, offering practical benefits for streamlining workflows and enhancing productivity within the SAP ecosystem. Discover more about Procurement & Agentic AI within this Spend Matters report