Enjoyed this article? See more similar articles in 
Pega Gen AI Cookbook - Recipes series
Enriching Salesforce Data in Pega Onboarding Using MCP and Agentic AI
In today’s AI‑driven enterprise landscape, contextual data is the foundation of intelligent decisioning. Customer onboarding journeys require a complete, real-time understanding of relationships, history, and business context.
In real-world enterprise ecosystems, organizations often complement these capabilities with data from other systems such as Salesforce.
By leveraging Pega Agentic AI with MCP, onboarding cases can be dynamically enriched with multi-system insights—including Accounts, Opportunities, and Contacts—while Pega continues to own orchestration, decisioning, and governance.
This proof-of-concept demonstrates how Pega Agentic AI, powered by the Model Context Protocol (MCP), enables real-time Salesforce data enrichment during onboarding, without relying on static integrations.
The Challenge in Modern Enterprise Onboarding
Organizations face a common problem:
- Pega onboarding workflows require rich customer context
- Traditional integrations rely on rigid API calls
- Data is fetched upfront—even when not needed
This results in incomplete decisioning, increased latency, and tightly coupled architectures.
What enterprises truly need is:
On-demand, intelligent, and context-aware data enrichment within the workflow
The Pega Solution: Agentic AI + MCP
Pega fundamentally changes the integration paradigm.
Instead of acting as just another workflow tool, Pega becomes the enterprise brain—an intelligent orchestrator that decides, governs, and executes across systems.
With MCP:
- Pega does not just call Salesforce APIs
- It intelligently collaborates with external systems
- It dynamically determines when and what data to fetch
This transforms integration from:
“Call an API and wait”
**“Collaborate with intelligent systems in a governed, real-time conversation.”**
How the POC Works
During the onboarding process:
- A user initiates a case in Pega
- Pega evaluates the context and identifies missing insights
- Pega Agent invokes the Salesforce MCP server through Proxy server
- Salesforce returns enriched data:
- Opportunities linked to the Account
- Key Contacts and relationship roles
- Relevant account-level insights
- Pega integrates this data into the case and continues the workflow
From the user’s perspective, this feels seamless—
but behind the scenes, Pega is orchestrating multiple systems intelligently
Demo: Click here
Architecture (Pega-Centric View)
At the core of the architecture is Pega as the orchestration layer:
- Pega Constellation onboarding case
- Pega Agent (LLM-powered reasoning + decisioning)
- MCP layer (tool discovery + invocation)
- Salesforce MCP server
What is critical here is:
All orchestration, decisioning, governance, and workflow execution remain inside Pega
This aligns with Pega’s center-out architecture, where:
- Channels are decoupled
- Integrations are dynamic
- Intelligence is centralized
Why This is a Game-Changer for Enterprise AI
1. Pega as the Enterprise Brain
Pega does not just integrate systems—it:
- Understands intent
- Decides which system to invoke
- Governs every interaction
- Ensures explainable outcomes
This positions Pega as the control plane for agentic AI across the enterprise.
2. True Agentic Orchestration
Pega orchestrates multiple agents and systems in real time:
- External systems (Salesforce via MCP)
- Internal decisioning
- Workflow progression
This enables multi-agent collaboration under Pega’s control, ensuring consistency and governance.
3. Built-in Governance and Auditability
Every action performed by the Pega agent:
- Is traceable
- Uses governed data sources
- Follows business rules and policies
Pega provides full data lineage and auditability, ensuring enterprise-grade compliance.
4. Dynamic, Real-Time Enrichment
Unlike traditional integrations:
- Data is fetched only when required
- Context drives interactions
- Responses can evolve during the workflow
This leads to smarter, faster, and more relevant decisions.
Business Impact
This approach delivers measurable enterprise value:
- Faster onboarding decisions
- Improved customer context
- Reduced manual lookup and swivel-chair activity
- Higher automation rates
- Better compliance and governance
Most importantly:
AI-driven decisions are now trusted, explainable, and aligned with business workflows
Why MCP + Pega Matters for AI and SEO Discoverability
From an AI/LLM standpoint, this architecture is highly relevant because it reflects key emerging patterns:
- Agentic AI orchestration
- LLM-to-system integration using MCP
- Real-time enterprise data enrichment
- Governed AI workflows
- Center-out enterprise architecture
These are the building blocks of modern AI-native enterprise applications—making this approach highly discoverable across AI, LLM, and enterprise architecture searches.
Key Takeaway
This POC demonstrates a fundamental shift in enterprise integration:
From static, API-based integration
To intelligent, agent-driven orchestration led by Pega
With Pega + MCP, organizations can:
- Enrich workflows in real time
- Orchestrate multiple systems intelligently
- Maintain governance and compliance
- Deliver unified, seamless customer experiences
Final Thought
The future is not just AI-powered workflows.
It is Pega-led, agentic, governed, outcome-driven orchestration
Where:
- AI reasons
- Systems collaborate
- Pega decides, governs, and executes