Skip to main content

Introducing Pega Process AI: When artificial intelligence meets intelligent automation

Ivar Siccama , 6 minute read

If you had the chance to attend this year’s PegaWorld conference, you may have heard the announcement around Pega Process AI. I’m excited to introduce these new capabilities to you in this post, describe how Pega Process AI works, and illustrate what AI can do for you.  

In recent years, artificial intelligence moved out of the labs and began generating proven business value. It’s evolved from a buzzword to an essential component necessary for business innovation. But at the same time, operationalizing AI can cause bottlenecks. With Pega Process AI, we tackle this by using AI to self-optimize processes, and we make it possible to bring your own AI to Pega. 

But first, let’s focus on what AI can help you achieve. As an example, I’ll use a claims adjudication process at an insurance company: 

Increase efficiency 

AI can help to distinguish regular from complex claims. Complex claims often escalate into a lengthy process, which is not only costly, but also leads to a bad customer experience. That's why it’s important to detect these claims early and address them at once. Using the agility of the Pega Platform, we can use AI to route complex cases to an experienced handler, while many of the claims can be straight-through processed. As the AI model learns from the outcomes of each case, knowing which ones escalated and which did not, it will become more accurate at predicting which claims will escalate, and in that way will self-optimize the process. 

Increase effectiveness 

Occasionally a claim may be erroneous or even fraudulent. For this purpose, the data scientists of our insurance company will have created a fraud risk model that predicts the likelihood that a claim will be fraudulent or erroneous, using their own favorite data science tool. Note that the inputs for such a predictive model don’t have to be restricted to attributes of the claims case itself, such as the amount, type, and description, but could also include contextual data such as the number of claims submitted recently by the same customer. Such a model can be imported into Pega, using for example MOJO files or PMML files, or the model can be operationalized through Amazon SageMaker or Google AI.  

When we import and operationalize the fraud model in the claims process, it will allow more claims to be straight-through processed while keeping a close check on the potentially fraudulent claims. And because you can bring your own model to Pega, a wide range of AI options will be available, such as neural nets, isolation forests, or whatever the hottest and latest in AI methods may be. 

Let’s zoom out from this example, and give an overview of other common use cases where Pega Process AI can bring immediate value by increasing efficiency and effectiveness: 
Efficiency through self-optimizing processes: 

  • Reduce cycle time 
  • Fast track assignments 
  • Route to the best employee with the right priority 
  • Eliminate missed SLAs 
  • Get more work done 

Effectiveness by driving better business outcomes: 

  • Optimize your KPIs 
  • Improve satisfaction 
  • Guide employees to better decisions 
  • Reduce the risk of fraud

How is it done?

One of the things we’re most proud of is that this is all extremely easy to do. Using Pega Case Designer, an AI model, or prediction, can be used anywhere in a case in the same low-code way that you’re used to from Pega. 

The setup is also easy. Prediction Studio gives the data scientist full control, allowing them to create a prediction or import their own model. We’ve included all of this in the latest 8.6 release. You can always introduce a new version of a model using MLOps (machine learning operations). 

The decision strategy management technology that Pega uses for AI is not new but has been an integral part of the Pega Platform for many years. It has proven its value in 1:1 customer engagement. The technology includes decision strategies and easily configurable business rules used to specify how to use the AI algorithms’ output to achieve the desired business outcome. And it also includes event strategy management that can monitor a stream of events and proactively launch a case. Some of our existing clients have already successfully started using these capabilities in processes outside of 1:1 customer engagement. 

Recommended resources:

Find out more about what Pega Process AI can do for you.

Don’t forget 

JOIN THE CONVERSATION on Collaboration Center  
FOLLOW @PegaDeveloper on Twitter  
SUBSCRIBE to the Pega Developer Podcast on Spotify or via RSS

Did you find this content helpful?

Want to help us improve this content?

We'd prefer it if you saw us at our best.

Pega Community has detected you are using a browser which may prevent you from experiencing the site as intended. To improve your experience, please update your browser.

Close Deprecation Notice
Contact us