We recently completed a POC connecting Amazon Quick to an existing Pega application through Pega Infinity '26’s MCP Service.
Amazon Quick, and other AI assistants, are already becoming the front door to enterprise knowledge for many organizations. Employees can use natural language to search across documents, systems, and data sources they’re already connected to, making it much easier to find the information they need. The last mile, however, is taking action with this information, and the action, oftentimes, is in Pega.
Using MCP, we enabled Amazon Quick users to initiate and complete governed business processes directly within an existing Pega application. Rather than building a separate AI experience, we exposed proven workflows, business rules, validations, and case management capabilities through a conversational interface.
Some key benefits of this approach:
Meet users in the AI assistant they already use every day
Leverage enterprise knowledge and context available through AI assistances
Turn insights into action through existing Pega workflows
Eliminate duplicate business logic and integrations
Preserve governance, auditability, and compliance controls
Future-proof AI investments through open standards like MCP
With this approach, organizations don’t need to choose between enterprise AI assistants and enterprise workflow. By combining the two, users can seamlessly move from finding information to executing work, all while leveraging the governance and process orchestration capabilities that already exist in Pega.
Excited to see how MCP continues to unlock new channels into Pega applications and accelerate adoption of enterprise AI!
Totally agree with meeting the users with where they are. This used to mean software like Outlook, Excel, or SAP, but is quickly become their AI interface of choice.
Would love to see a demo of the POC! - Feel free to share any further detailed info in this regard.
I’m curious to know more about the authentication and authorization mechanism connecting external AI assistants to Pega Infinity '26’s MCP Service to access the application in a secure manner to handle multiple user variants would like to access MCP services.
Thanks, Krishnakumar! I will share the demo here shortly. To answer your security question, Pega Infinity ’26 MCP Service provides a standards-based, secure integration layer for external AI assistants. User authentication is handled through OAuth-based identity flows, while authorization is enforced by Pega using the application’s existing security model. This ensures that AI-driven interactions inherit the same role-based access controls, governance policies, and audit capabilities already defined within the enterprise application, regardless of which AI assistant or channel initiates the request.