Why You Need Both RPA & AI in Your Procurement Processes

If you’ve spent any time in procurement circles lately, you’ve probably heard the terms robotic process automation (RPA) and artificial intelligence (AI) thrown around like they’re interchangeable. Spoiler: They’re not. Understanding how they differ, and more importantly how they work together, is key to building smarter, more resilient procurement processes.

Let’s dive in!

What is robotic process automation & how does it support procurement?

As the name suggests, RPA is a robot or “bot” you can program to complete repetitive, rules-based tasks like data extraction and data entry. Give it specific parameters and instructions and RPA will follow those instructions exactly. Once you tell it what to do, RPA is incredibly hard working — it never sleeps, takes breaks or goes on vacation. It can also click, read and type far faster than a human.

However, RPA software isn’t very smart. It performs best in processes that are rigid, highly-structured and don’t change much. When RPA encounters something outside the parameters of its programming, it’s unable to complete the task. Any time an occasional variable crops up in your process (or you adjust something in your process), you have to retrain your bot so it knows how to handle those changes.

Examples of how RPA is used in procurement:

• Sending notifications whenever a contract is approaching its expiration date
• Approving routine purchase requests that fall within a company’s procurement strategy
• Conducting 3-way matching, comparing purchase requests with supplier invoices and delivery receipts
• Extracting info from suppliers’ websites, such as product lists, prices, references, etc.

What is artificial intelligence & how does it support procurement?

AI technology can analyze large datasets to learn, make decisions and solve problems. It can take on complex tasks such as speech recognition, image analysis and language translation. It can even make predictions and recommendations based on what it’s learned.

AI performs best when it has large amounts of high-quality data to draw from. The more information it has, the more it can learn and improve. In other words, AI can be taught. Give AI a clear definition of what “success” or “good” looks like in a process and it can quickly follow your example to complete the process the way you want. And if it encounters variables or things that don’t fit into what you told it was “success” or “good,” it’s smart enough to bring it to your attention so you can give it direction. In addition to flagging any anomalies for human review, AI can also recommend how to correct them.

Examples of how AI is used in procurement:

• Improving spend classification accuracy
• Consolidating data from multiple sources to paint a complete picture of supplier performance and proactively identify potential supplier disruptions
• Sifting through mountains of data to uncover hidden cost-saving and revenue-generating opportunities
• Delivering data-driven, ready-to-push-go recommendations

What’s the difference between RPA & AI?

Think of RPA like a macro in a spreadsheet, executing a series of specific commands such as “go to this cell” or “do this action.” If anything in the spreadsheet configuration changes (for example, a column or cell moves) the macro breaks and can’t complete its actions. Similarly, RPA can’t function if something changes in the process it’s been programmed to complete. It’s like a brand new employee who needs very clear instructions if you want it to do something.

Meanwhile, AI is like a seasoned supervisor who the RPA can go to when something doesn’t work as expected. It’s a higher order intelligence that can look at situations that contain some variance and tell the RPA what it needs to do differently. It helps RPA keep working when something goes wrong without as much human intervention needed. And if AI does need a human to give it direction, it will proactively surface any issues to its human user — no need to remember to go in and reprogram your RPA anytime a change occurs. 

Why you need both RPA & AI in procurement

This isn’t an either/or situation, it’s a both/and situation. You want both RPA and AI (often called intelligent automation or AI-driven automation) in your procurement processes. In fact, AI already has RPA built into it. Because AI can understand how to identify a variance in a process, it already understands how to do the repetitive tasks in that process. It can do everything RPA does, but with more thought processes and decision-making capabilities.

RPA by itself isn’t bad, it just only gets you so far. If you’re in an environment where things don’t change much, maybe RPA is enough for you. But as our world keeps evolving and becoming more complex, those environments are becoming less and less common. Because AI includes RPA, it can automate your tasks but also identify issues and variances without requiring constant re-programming.

How RPA & AI work together in procurement

The RPA/AI combo can do some pretty neat things in the procurement process. Here are some examples of how Esker Procurement, an AI-driven automation system, leverages intelligent automation to make procurement faster and easier.

Anomaly detection in purchase requisitions

Manually checking every single incoming purchase requisition to catch quantity errors takes FOREVER. Esker Synergy AI within the Esker Procurement platform analyzes your historical order data so it knows what your organization typically purchases. Then, when it detects unusual product quantities or purchase amounts, it sends automated alerts to the appropriate stakeholders so a human can investigate the purchase requisitions in question.

Intelligent automation drastically reduces the time and effort your team needs to spend on order validation while preventing overstocking and shortages that can result in overspending and operational disruptions.

Intelligent data extraction from supplier quotes

Turning supplier quotations into purchase requisitions is an annoying bottleneck in the procurement process. Or at least it was until Esker Synergy AI came along. It automatically converts supplier quotations into purchase requisitions by reading and extracting the necessary info from quotes and auto-populating it into the correct requisition fields. AI-driven automation eliminates time-consuming manual data entry, reduces errors that can cost your company money and speeds up the procurement cycle.

Getting started with RPA & AI in your procurement process

Bringing the combined benefits of RPA and AI to your organization doesn’t have to be complicated, confusing or overly costly. Esker’s AI-powered procurement software is a cloud-based application that seamlessly connects to just about any ERP — even if your company uses multiple different ERPs. It offers:

• A clean, simple user interface that requires minimal training to learn
• Customizable dashboards that display the metrics that matter to you and can quickly generate reports
• A mobile app that lets managers approve purchase requisitions and review KPIs and metrics anytime, anywhere
• A supplier self-service portal that streamlines catalog management, gives access to invoice payment status 24/7 and facilitates fast resolution of exceptions and discrepancies

And because Esker Procurement is part of our complete platform designed to automate your entire source-to-pay and order-to-cash cycles, it’s easy to gradually implement AI-driven automation into more areas of your financial processes as your team gains confidence and you start seeing results.

Why not see for yourself what your procurement process could look like with Esker by watching our on-demand demo? 

​