New capabilities connect third-party agents to Pega processes, delivering predictable outcomes with predictable costs
Pegasystems Inc. (NASDAQ: PEGA), the enterprise AI software company for mission-critical work, today at PegaWorld® announced new capabilities that enable enterprises to use their custom-built AI agents to drive essential business processes without sacrificing governance, compliance, or cost controls.
This press release features multimedia. View the full release here: https://www.businesswire.com/news/home/20260608601073/en/

This screenshot shows how every application built on Pega is now a Model Context Protocol (MCP) server—enabling your predictable Pega workflows to be used in your AI agent of choice.
With added support for the open MCP (Model Context Protocol) standard, Pega processes can now be discovered and executed by authorized third-party agents, including those built on Anthropic Claude, Google Gemini, OpenAI, AWS AgentCore, and more. This opens the growing universe of agents to millions of Pega-powered workflows, giving organizations predictable outcomes and predictable costs for their mission-critical work.
Market Context: More agents, more uncertainty
AI agents hold enormous promise, but hallucinations, compliance gaps, and runaway token costs are stalling projects before they can deliver value. Gartner reports over 40% of agentic AI projects will be canceled by the end of 2027 due to escalating costs, unclear business value, and inadequate controls. Without a smart approach to reliably orchestrate and run their agents, enterprises risk turning agent innovation into operational uncertainty.
A Closer Look: Predictability, speed, and scale where it matters most
With Pega’s built-in MCP capabilities, enterprises can deploy agents while maintaining the reliability, auditability, and predictability needed to run their core processes. Other approaches typically require agents to re-reason the process at every step, risking unpredictable results and huge token costs. By connecting to Pega’s industry-leading business orchestration and automation technology (BOAT) platform, agents can discover and execute Pega workflows, which then guide the agent step-by-step for consistent and predictable outcomes and cost.
At each workflow step, Pega can invoke external agents to automate manual work and reuse agents built for other purposes. Pega Blueprint AI™ workflow design agent helps identify where these step agents can increase efficiency, while Pega’s industry-leading orchestration audits agent interactions across the process lifecycle.
In addition, Pega introduced pre-built agents available to help drive further automation:
- Proactively engage a person when needed: The new agentic assignment agent uses AI to reach out to an employee or customer when additional data or approvals are needed to complete a request. This approach pulls in a human via email, chat, or telephony without requiring manual outreach.
- Automate document processing: The new document agent analyzes and acts on complex documents – intelligently splitting, categorizing, bounding, scoring, and triaging them for further processing. A new chat feature lets employees ask questions about their PDFs, documents, and images for instant insights.
Availability
The MCP server capabilities, agentic assignments, and document agent will be available to Pega clients as part of the latest Pega Infinity™ 26 release expected in Q3.
The solutions are on display at PegaWorld, the annual user conference at the MGM Grand in Las Vegas being held this week. Watch Pega Chief Product Officer Kerim Akgonul’s keynote address with demonstrations of Pega Infinity 26 capabilities at approximately 10:15 am Pacific Time on Monday, June 8 at www.pegaworld.com or watch the replay later.
Quotes & Commentary
“Agentic AI is moving fast, but without a reliable way to run, scale, and orchestrate agents, enterprises risk trading innovation for instability,” said Kerim Akgonul, chief product officer, Pega. “Pega’s new MCP capabilities give organizations an easy way to connect their AI agents to their mission critical processes to orchestrate predictable outcomes with predictable cost.”
“As enterprises accelerate adoption of AI agents, the real challenge is executing with control and predictability at scale,” said Deepak Tiwari, global Pega leader, EY. “Pega’s support for MCP further strengthens its agentic orchestration capabilities, enabling organizations to operationalize and scale AI within mission-critical processes. At EY, we see this as a key enabler for driving measurable outcomes via AI with enterprise-grade governance.”
"Pega's Model Context Protocol support is a breakthrough for the agentic AI ecosystem,” said SrinivasaRao Ganta, global head, digital process management, Pega practice, Tata Consultancy Services. “For the first time, our AI agents can invoke enterprise-grade Pega workflows — think dispute resolution, claims processing, or customer onboarding — directly and securely, without building custom adapters or point-to-point integrations. The fact that this works across any third-party agent framework, backed by enterprise-ready security and seamlessly connected to Pega Blueprint designs, means we can deliver truly interoperable agentic solutions at enterprise scale. This is the connective tissue the industry has needed."
Supporting Resources
- Learn more: Pega agentic orchestration
- Token Cost Calculator: See how much you are wasting in LLM costs
- Background: Pega Predictable AI™ Agents
About Pega
Pega delivers the platform to reimagine, run, and evolve the processes and decisions an enterprise can't afford to get wrong. We combine AI with proven architecture to keep mission-critical operations governed, scalable, and continuously adaptable. Since 1983, the world's largest organizations have trusted Pega to turn transformation ambition into durable results. Learn more at pega.com.
All trademarks are the property of their respective owners.
View source version on businesswire.com: https://www.businesswire.com/news/home/20260608601073/en/
With Pega’s built-in MCP capabilities, enterprises can deploy agents while maintaining the reliability, auditability, and predictability needed to run their core processes.
Contacts
Press Contact:
Sean Audet
Pegasystems
sean.audet@pega.com
Pega on LinkedIn
If you believe this article contains misleading, harmful, or spam content, please let us know.
Report this article