Skip to main content

How can I tell if my use case is Process AI or Pega CDH?

Joe Carew, 6 minute read

Pega has been a market leader in the use of Artificial Intelligence (AI) to support marketing automation programs using Pega CDH. Given the power of this capability, it is unsurprising that many organizations have asked how this same AI engine could be used to support use cases that go beyond marketing and customer engagement. In response to increasing demand from our users, in 2021 we released Pega Process AI, which allows Pega Platform clients to enhance their experiences with decisioning, event streaming, and AI functionality.

Organizations that have Pega CDH and Pega Platform may wonder about where AI decisioning in Pega CDH ends and Pega Process AI begins. In this post, I’ll briefly describe Pega Process AI and then compare and contrast it with Pega CDH so that Artificial Intelligence can be leveraged across the enterprise in ways that improve ease of use, efficiency, and impact.

What is Process AI?

Pega Process AI lets businesses supercharge their workflows, combining the power of time-saving automations with “smart,” self-optimizing AI to drive real results at scale.

Process AI is the only process automation engine with enterprise-scale machine learning predictions, event processing, NLP, and decisioning built in. With Process AI, you get a solution that combines case management, back-end integrations, automation, and AI seamlessly – all in a low-code platform that accelerates time-to market. That means, instead of waiting months to see results, you can start harnessing your data to optimize your processes instantly.   

Typical applications of Process AI in a workflow include using a Prediction to determine things like:

  • Which path in a process flow is most likely to result in the case meeting its SLA?
  • Which work queue is most likely to reduce the cost of the case?
  • Is the case likely to be fraudulent?

Process AI is especially well-suited to event handling, which involves taking some action (i.e. initiating some process) based on a defined set of triggers. An example of event handling is receiving events from an industrial machine, running an event strategy to aggregate the machine’s telemetry, and emitting an event trigger that results in a service case being created for the local dealer.

When to use Pega Process AI instead of Pega CDH

A seasoned implementor with Pega Decisioning experience already knows that if an organization is licensed for it, they can call CDH Real-Time APIs in a process flow to get Next Best Actions (NBAs) to use in a case.

But this is not the best approach when the decision is about how to route work or which direction to go in a workflow.  

Pega Process AI removes the customer-centric outputs of CDH from the mix and provides integration directly within case types, allowing you to easily operationalize decisions that factor in the work being done.

An example of a Process AI implementation is using a Prediction in a process flow that gives the system a probability (likelihood) to use in deciding whether to go down Flow Action (path) A or B. Pega Process AI allows you to use Predictions created in Prediction Studio and configure a case to use that Prediction.  The probability determined by the model can be used in a conditional ‘When’ component configured using our simple Condition Builder in App Studio.  More advanced approaches, from executing a strategy in a utility shape to integrating decisions based on case data, are also supported by Process AI.

Here are some things to consider when trying to decide whether to use Pega Process AI or Pega CDH to power your low-code user case:

Attributes of a Process AI use case

  • Established workflows that leverage Pega's core case management capabilities and have enough volume to be able to "learn" from case history—typically not an issue--or enough data from other sources to support the real-time analytic they wish to embed.
  • Reportable KPI and/or business value for the workflow
  • The workflow includes properties that are not centered primarily on customer interactions; complex transactions, agents, machines and other entities are in the mix
  • The proposed Prediction is that of a property used in a process or the probability of an outcome or set of outcomes
  • The available properties—derived or fetched by the process--are good predictors. For example, absolute values like dates, timestamps, financial values, identifiers, and Personally Identifiable Information (PII)
  • Events being handled result in workflows being created or updated

Signs pointing to a CDH use case

  • The decision involves how or what to communicate to a customer
  • The primary predictors being used in decisioning come from customer-related data objects
  • There are complex eligibility/engagement policies and arbitration/ranking rules that need to be maintainable by business users


Pega has an experienced team of specialist Solution Consultants called Advanced Platform Specialists that is prepared to help customers and integration partners understand how and when to apply Process AI to optimize workflows. If you have questions or are considering a non-standard scenario and need help to determine the best way to create a truly intelligent workflow, ask your Pega account team to contact the Advanced Platform Specialist team for demonstrations, Q&A, overviews, and clinics.


Learn More 

Don't forget 

About the Author

Joe Carew is a principal solutions consultant at Pegasystems 

Share this page Share via x Share via LinkedIn Copying...

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