Revolutionising Procurement: A Deep Dive into AI Agents

Read this if you're serious about Procurement

Feb 17, 2025

Hey Procurement Legend,

I’ve been busy putting some thoughts together on Agentic AI in Procurement. Here’s the start of this new series where we’ll explore the weird, the wonderful, and get you up to speed with what could happen.

If you find it valuable, please consider becoming a paid member. I’m releasing this as a “Free” article for anyone to check out, and all future articles will be for paid members only. It’s the price of two coffees a month and will accelerate your procurement career. Also, you should be able to claim this as a business expense with your employer.

Procurement has long been viewed as a back-office function bogged down by manual tasks, legacy systems, and inefficient processes. Many of you, including me, have had to deal with the issues. Today, however, artificial intelligence (AI) is transforming procurement. At the heart of this change are AI agents—also known as agentic AI—autonomous software entities that execute tasks with minimal human oversight. In this article, we’ll explore what AI agents are, the various forms they take, and how they are poised to revolutionise procurement across organisations of all sizes.

What Is an AI Agent (Agentic AI)?

AI agents represent a fundamental shift in automation. Unlike traditional scripts that follow fixed instructions, AI agents learn, adapt, and make data-based decisions. They are not merely “if-then” tools; they actively interpret complex environments and dynamically adjust their actions. Here’s a closer look at what defines an AI agent and the broad categories they fall into:

Key Characteristics of AI Agents

Data Perception and Interpretation:
AI agents continuously gather data from their environment through sensors, digital inputs, or online data streams. Using technologies like natural language processing (NLP) or computer vision, they interpret this information to understand context and meaning. This capability allows them to “see” patterns and nuances that a static program cannot.
Planning and Decision-Making:
Advanced AI agents don’t just react; they plan. They analyse available data to determine the best action, often evaluating multiple options and predicting potential outcomes. This means they can adjust their strategy as circumstances change—a key advantage over rigid automation.
Autonomous Action:
Once an AI agent makes a decision, it acts autonomously. Whether sending an automated reply, executing a command, or coordinating with other systems, the agent carries out its tasks without waiting for further human instruction.
Learning and Adaptation:
Through methods like reinforcement learning, where actions are rewarded or penalised, AI agents learn from their experiences. Over time, they refine their decision-making processes, improving accuracy and efficiency. This continuous learning makes them increasingly valuable as they adapt to new data and scenarios.

Broad Categories of AI Agents

AI agents come in different flavours, each suited to particular tasks and environments. Below are some of the most common categories:

Reactive Agents:
These agents operate on a simple stimulus-response basis. They don’t store past interactions or plans—they simply react to incoming data. An example is a basic chatbot that answers frequently asked questions. While highly efficient for routine tasks, reactive agents lack the flexibility for more complex operations.
Proactive Agents:
Proactive agents take a step further by predicting future needs and planning accordingly. They can assess multiple scenarios and choose an action based on what they forecast will be most beneficial. This ability makes them suitable for environments where anticipating changes is critical, such as real-time risk management.
Hybrid Agents:
Combining the strengths of reactive and proactive agents, hybrid agents can quickly respond to immediate stimuli while using historical data to inform longer-term strategies. This balance allows them to adapt to dynamic conditions without sacrificing speed or foresight.
Utility-Based Agents:
These agents evaluate possible actions against a utility function—a mathematical measure of the desired outcome. They select the option that maximises this utility. Such agents are particularly useful for optimisation tasks like resource allocation or route planning, where the goal is to achieve the best possible result.
Learning Agents:
Learning agents continuously improve by analysing the outcomes of their actions. Using machine learning techniques, they adjust their strategies based on feedback, becoming more effective over time. This category includes agents that use reinforcement learning to fine-tune their decisions based on rewards and penalties.
Collaborative Agents:
A single agent may not have all the required expertise in many complex tasks. Collaborative or multi-agent systems consist of multiple specialised agents that work together to achieve a common goal. They communicate and coordinate, ensuring that the overall system is greater than the sum of its parts.

Applications Beyond Procurement

While this article focuses on procurement, AI agents are already transforming many other fields. For instance:

Buyer Agents (Shopping Bots): Automate online searches for the best deals.
Personal Agents (User Assistants): Manage schedules, emails, and other tasks (think Siri or Alexa but wayyyyyy smarter because SIRI still sucks).
Monitoring Agents: Monitor system performance, flag anomalies, or even help in cybersecurity.
Data-mining agents: Analyse vast datasets to uncover trends and insights crucial for market analysis and predictive maintenance.
Software Development and Security Agents: Automate code reviews or monitor networks for security breaches.

From these diverse applications, it’s clear that AI agents have matured far beyond simple automation. Their capabilities to perceive, plan, act, and learn set the stage for revolutionary changes across all sectors, including procurement.

Transforming Procurement with AI Agents

The promise of AI agents is particularly compelling in procurement—a function that manual, repetitive tasks and legacy processes have long characterised. The most significant opportunities for AI in this space lie in transforming the way procurement teams work by automating tasks that have traditionally been error-prone or neglected. This transformation unfolds through three key avenues:

1. Domain-Specific Workflows

Every procurement process—from processing purchase orders and managing contracts to conducting supplier performance reviews and analysing spend data—has its own set of unique challenges. Despite being slow and inefficient, these tasks have long been viewed as “just the way things are.” Despite attempts to digitalise them, many still feel there’s too much noise to get them working.

Automating Complex Processes

Take purchase order processing as an example. Traditionally, this task involves multiple steps: manual data entry, verification across various systems, and cross-checking for errors. A mistake at any stage can lead to delays, compliance risks, and higher costs. AI agents can automate these tasks by extracting relevant data, validating it against internal databases, and flagging discrepancies. This not only speeds up the process but also reduces human error.

Similarly, contract management is another area where AI can add immense value. AI agents with natural language processing capabilities can scan lengthy contracts to extract key details—renewal dates, critical clauses, and risk indicators—providing procurement teams with actionable insights.

Unlocking New Areas for Efficiency

By automating these domain-specific workflows, organisations free up resources to focus on strategic activities like supplier relationship management, market analysis, and strategic sourcing. This reallocation not only cuts costs but also opens up opportunities for innovation. For example, a study highlighted by Sievo shows that companies leveraging AI in procurement can make more informed sourcing decisions through timely analytics and data-driven insights. This means potential savings and competitive advantages that were previously out of reach.

2. Empowering SMEs and Resource-Constrained Teams

Not every organisation has the extensive procurement resources that large enterprises enjoy. Small and mid-sized enterprises (SMEs) often face the same complex challenges, with fewer dedicated personnel and less sophisticated systems. AI agents offer these organisations a chance to level the playing field.

Augmenting Limited Resources

Automating routine tasks like order tracking, invoice processing, and supplier negotiations can transform SMEs. AI agents are an instant extension of the procurement team, handling day-to-day operations that would otherwise require significant workforce. This reduces operational costs and enables SMEs to scale their operations quickly.

Startups and smaller businesses can leverage AI agents to streamline their procurement processes. Automating data collection and spend analytics allows these companies to quickly identify the best suppliers and negotiate favourable terms, minimising the risk of manual errors. Procurement Magazine has showcased several AI tools designed to empower smaller procurement teams and enable them to compete with larger organisations.

Fostering Growth and Innovation

By taking over repetitive tasks, AI agents free business leaders to focus on strategic growth. This extra bandwidth can be directed toward exploring new market opportunities, developing innovative products, or enhancing customer service. The agility provided by AI agents means that SMEs can adapt quickly to market changes—finding alternative suppliers during disruptions or capitalising on emerging trends. In an increasingly competitive landscape, this nimbleness is crucial for long-term success.

3. Advanced Orchestration & Strategic Intelligence

In large enterprises, procurement is evolving from a reactive, back-office function to a dynamic, strategic operation that drives competitive advantage. AI agents in this context do more than automate tasks—they redefine procurement strategy.

Building a Network of Specialised Agents (what I’m most excited about)

Imagine a network of specialised AI agents that collaborates across your entire procurement ecosystem. These agents monitor global supplier performance, simulate multiple sourcing scenarios in real time, and even predict market disruptions before they happen. By integrating seamlessly with ERP systems and digital twin platforms, these agents blend real-time data with historical trends, offering procurement teams a comprehensive view of their supply chain.

A Wall Street Journal report has detailed how companies like Salesforce and ServiceNow are already investing in AI agents that provide strategic insights—extending their role from simple task automation to proactive decision-making.

Proactive Strategy and Risk Mitigation

Advanced orchestration transforms procurement into a proactive function. AI agents can simulate different sourcing strategies and predict outcomes, allowing procurement teams to lock in favourable terms or explore alternative options before disruptions occur. By factoring in variables like supplier reliability, geopolitical risks, and market trends, these agents help develop strategies that minimise risk and maximise value.

For instance, during periods of market volatility, AI agents can analyse historical pricing data to forecast future trends. This predictive capability enables procurement teams to make informed decisions quickly—turning potential risks into strategic opportunities. As industry analysts note, organisations that successfully integrate AI into their procurement processes can gain a significant competitive advantage through improved agility and strategic foresight.

Turning Procurement into a Strategic Powerhouse

At its core, advanced orchestration through AI agents transforms procurement into a competitive advantage. Large enterprises can optimise costs and drive innovation with real-time dashboards, predictive analytics, and automated decision-making. This shift enables procurement to move from being a cost centre to a strategic driver of growth and innovation.

If your enterprise is ready to move beyond traditional procurement methods, advanced orchestration via AI agents could catalyse a complete paradigm shift. These agents can transform procurement from a reactive, manual process into a dynamic, strategic powerhouse that reduces costs, fosters innovation, and minimises risks.

Integrating AI Agents into the Procurement Ecosystem

While the benefits of AI agents are clear, their successful integration requires careful planning and execution. Here are some key considerations:

Data Quality and Integration

AI agents rely on high-quality, up-to-date data. Ensuring that procurement data is accurate and consolidated from various sources—such as ERP systems, supplier databases, and contract management systems—is essential. As experts stress, robust data integration forms the foundation of any successful AI deployment. However, using AI to extract and cleanse data could be the way to do this.

Change Management and Training

Adopting AI in a traditional procurement environment calls for a cultural shift. Teams must be trained to use new tools and interpret the insights these tools generate. Emphasise that AI agents are meant to augment human capabilities rather than replace them, creating a collaborative environment where technology and expertise work hand in hand. However, we must be realists. This level of AI can potentially displace roles en mass - in any profession.

Pilot Programs and Scalability

Start small with pilot programs to test the impact of AI agents on specific procurement functions. Once these pilots prove successful, scale the solutions gradually across the organisation. This incremental approach minimises risk and builds confidence in AI-driven processes.

Governance, Risk, and Ethical Considerations

As AI agents take on more complex roles, establishing transparent governance and risk management practices is critical. Define guidelines for data usage, decision-making authority, and accountability. Additionally, ethical considerations should be addressed to maintain trust and compliance, such as transparency in automated decisions and mitigating biases.

The Future of Procurement with AI Agents

The journey toward AI-driven procurement is already underway, and the potential is vast. Here’s what we can expect:

Dynamic Supplier Networks: AI agents could enable real-time adjustments to supplier networks, ensuring that procurement teams always work with the best partners.
End-to-End Automation: Future AI agents may manage entire procurement cycles—from order placement to payment processing—with minimal human intervention.
Proactive Market Intelligence: Advanced AI systems will provide continuous market intelligence, alerting procurement teams to emerging trends, disruptions, or new opportunities before they become apparent.
Strategic Decision Support: As AI agents become more sophisticated, they will serve as strategic decision-support tools that simulate various scenarios and forecast outcomes, driving long-term business value.

The potential of AI agents in procurement is not a distant dream—it’s a reality that is already reshaping industries. By automating manual tasks and providing strategic insights, AI agents offer a pathway to transform procurement from a necessary expense into a dynamic driver of growth and innovation.

Conclusion

AI agents—agentic AI—are revolutionising the automation landscape. They are more than simple rule-following scripts; they are intelligent, adaptive systems capable of perceiving data, planning actions, learning from feedback, and collaborating with other agents. Whether through reactive, proactive, hybrid, utility-based, learning, or collaborative models, these agents are set to transform how we approach everything, from everyday tasks to strategic decision-making.

The potential in procurement is enormous. AI agents can turn a traditionally manual process into a lean, agile, and competitive function by automating domain-specific workflows, empowering resource-constrained SMEs, and enabling advanced orchestration and strategic intelligence in large enterprises.

For organisations ready to move beyond outdated procurement practices, embracing AI agents is not just an option—it’s a strategic imperative. As the future unfolds, those who invest in AI-driven procurement today will unlock substantial cost savings, drive innovation, and secure a competitive edge in an increasingly dynamic global market.

I want to add a note here. I’m currently in the middle of this from a technology provider perspective. This is all relatively new. It’s going to be disruptive, but it will truly augment you.

I’m excited about it.

The AI Procurement Blueprint

Discussion about this post