Enjoyed this article? See more similar articles in 
Pega Gen AI Cookbook - Recipes series.
Enterprise AI is rapidly evolving into a new paradigm. Organizations are no longer focused on integrating a single AI model; instead, they are orchestrating multiple enterprise systems along with specialized AI agents to address increasingly complex business challenges.
Customer data resides within CRM platforms, while business processes execute within enterprise applications. AI models contribute advanced reasoning and recommendations. The core challenge is how to bring these capabilities together seamlessly—without creating tightly coupled integrations or compromising governance.
This is where Pega delivers significant value.
With the Pega MCP Connector, Pega acts as the Process Agentic Fabric—an orchestration layer that connects enterprise applications, AI agents, and business workflows into a unified and governed experience.
Rather than serving as just another integration point, Pega coordinates the end-to-end business process, enabling enterprise systems to contribute contextual data while AI agents provide specialized intelligence.
Pega Process Agentic Fabric
The Pega Intelligent Agent Fabric follows a structured and effective architectural approach:
- Pega manages the business workflow and overall customer journey
- Enterprise systems (such as Salesforce) provide enriched business context
- Enterprise LLMs deliver intelligent reasoning and recommendations
- Pega consolidates these inputs, applying business rules, governance, and human approvals before making the final decision
This pattern ensures that business ownership remains within Pega, while organizations continue to leverage best-of-breed enterprise systems and AI capabilities.
Demo: Click here
Demo Scenario
To demonstrate this pattern, a business loan application scenario was implemented.
The process begins when a customer submits a loan request through a Pega application.
Pega handles the complete intake process—capturing applicant information, validating required inputs, and initiating the case lifecycle.
Once intake is complete, Pega uses the MCP Connector to interact with Salesforce.
Instead of relying on direct point-to-point APIs, Pega leverages MCP to retrieve key CRM insights, including:
- Customer account details
- Existing opportunities
- Relationship history
- Previous interactions
This additional context enriches the application before any AI-driven evaluation occurs.
Pega then combines:
- Intake data captured within Pega
- Customer insights retrieved from Salesforce
into a single enriched business request.
Using the same MCP framework, Pega routes this request to Claude for intelligent loan assessment.
Claude evaluates the complete context and returns:
- Lending recommendations
- Risk assessments
- Supporting rationale
- Identification of missing information
- Suggested next actions
Finally, Pega applies enterprise business rules, decision strategies, routing logic, governance controls, and human approvals to determine whether the application should be approved, declined, or escalated for manual review.
Throughout the process, control of the workflow remains fully within Pega.
Why This Pattern Matters
Many AI implementations focus only on connecting applications to an LLM.
The Pega Intelligent Agent Fabric represents a broader and more enterprise-ready approach.
Rather than relying on AI in isolation, Pega orchestrates multiple capabilities:
- Enterprise workflows
- CRM systems
- Business context
- AI-driven reasoning
- Governance frameworks
- Human decision-making
Each component contributes its unique strength:
- Salesforce enriches the process with customer context
- Enterprise LLMs provide intelligent analysis
- Pega orchestrates and governs the entire workflow
Business Benefits
This architecture brings several advantages to enterprise organizations:
- Centralized orchestration of business workflows
- Seamless collaboration with enterprise systems through MCP
- Reusable and scalable AI integrations
- Strong governance and auditability
- Human-in-the-loop decision-making
- Reduced integration complexity
- Model-agnostic architecture
- Scalable and future-ready AI foundation
Most importantly, organizations can adopt new AI capabilities without redesigning their existing business workflows.
Conclusion
Enterprise AI is no longer about connecting a single model to one application.
It is about orchestrating enterprise systems, business processes, and specialized AI capabilities into a cohesive and governed experience.
The Pega Intelligent Agent Fabric demonstrates how Pega fulfills this role as an orchestration layer.
In this scenario, Pega manages the end-to-end loan workflow, enriches the application with Salesforce data via the MCP Connector, collaborates with Claude for intelligent evaluation, and ultimately governs the decision using its native workflow and decisioning capabilities.
While demonstrated through a loan application use case, this same pattern can be applied to claims processing, customer service, underwriting, onboarding, case management, and many other enterprise scenarios.
As organizations continue to adopt enterprise AI, the ability to orchestrate—rather than simply integrate—will become a key differentiator.
Salesforce provides context. Enterprise LLMs provide intelligence. Pega brings everything together.
Enjoyed this article? See more similar articles in Pega Cookbook - Recipes series.