Cognizant just made Workflow Agentic Processing generally available for TriZetto Facets and QNXT — alongside Enterprise MCP Servers and a library of 100+ MCP tools. The immediate use case is automating eligible routine pended claims. The bigger signal is that AI agents are moving directly into the payer core where claims, eligibility, prior authorization, and payment workflows are administered.
Cognizant announced general availability of two connected capabilities for the TriZetto Facets and QNXT core administration platforms: Workflow Agentic Processing and Enterprise Model Context Protocol (MCP) Servers.
Workflow Agentic Processing lets AI agents execute a health plan's standard operating procedures against eligible routine pended claims. When a claim requires judgment — a denial, exception, or complex case — it can be routed to a human reviewer instead of being forced through an autonomous path.
The MCP layer is what makes this more strategically important. Cognizant says health plans can use a library of more than 100 MCP tools to give their own AI agents access to TriZetto data and workflows, including use cases around member and provider service and prior authorization.
MCP is moving from an AI-developer convenience into core healthcare transaction infrastructure. When Facets and QNXT expose governed payer workflows to agents, the payer operating system itself becomes programmable by AI.
Pended claims sit in the middle ground between fully automated adjudication and work that truly needs a human. They are repetitive, rules-heavy, and expensive when they accumulate in queues — which makes them an obvious target for agentic automation.
Cognizant's design is notable because it does not position the agent as a black box. The system is built around health-plan operating procedures, with work queues, agent actions, and outcomes visible in the workflow. Organizations can adjust procedures as policies change rather than rebuilding automation logic from scratch.
That is closer to how enterprise RCM and payer operations will likely adopt agents at scale: bounded autonomy, explicit procedures, audit trails, and human escalation — not a general-purpose model making unconstrained claims decisions.
Most provider-side AI companies have focused on the revenue cycle from the outside: eligibility checks, coding, claim submission, status, denials, appeals, and patient collections. This launch matters because the payer side is automating the systems those provider workflows ultimately interact with.
TriZetto platforms support more than 200 million members and process more than $500 billion in annual healthcare spend across claims, eligibility, prior authorization, and payment integrity. Putting agentic workflows into that layer could affect the speed and consistency of payer operations at enormous scale.
For providers, the long-term implication is straightforward: the counterparty to your revenue cycle is becoming more automated too.
Provider organizations should expect more machine-to-machine interaction with payers. Structured documentation, cleaner claim data, API-ready workflows, and evidence that can be interpreted reliably by automated payer systems will matter more — not less — as adjudication and exception handling become increasingly agent-driven.
Workflow Agentic Processing is the product. MCP is the platform strategy.
By exposing TriZetto capabilities through a governed MCP tool library, Cognizant is making it easier for health plans to connect their own AI initiatives to core administrative data without building a bespoke integration for every new agent. That creates a common interface between AI systems and payer operations.
This is the same architecture shift showing up across enterprise software: the valuable system is no longer just the application UI. It is the tool layer that allows agents to safely retrieve context and take actions.
In healthcare, that matters because the underlying workflows are unusually fragmented and regulated. If MCP becomes a practical standard for agent access to claims, eligibility, prior auth, and payment data, the next generation of healthcare automation could be built around interoperable agent tools rather than point-to-point bots.
The release is commercially meaningful, but it is still early. Cognizant says early-adopter health plans are piloting Workflow Agentic Processing and that performance results will be shared later. There are not yet public metrics showing claim-level accuracy, auto-resolution rates, cost savings, or downstream effects on denials and provider payment timing.
So the strongest conclusion today is architectural, not performance-based: Cognizant has moved agentic AI and MCP into production-grade payer core administration. The operational ROI still needs to be demonstrated.
The revenue cycle AI race is no longer happening only on the provider side. Payers are beginning to deploy agents inside the systems that adjudicate claims and administer benefits.
That creates a new operating environment where provider agents may increasingly interact with payer agents — both working from structured policies, transaction data, and governed toolsets. The advantage will shift toward organizations that can make their revenue cycle data cleaner, their evidence more structured, and their workflows more machine-readable.
Cognizant's TriZetto launch is one of the clearest signals yet that agentic RCM is becoming infrastructure, not a feature.
Deal signals, payer shifts, and AI developments — before they reshape your revenue cycle. Free, every week.
Source: Cognizant announcement · RevCycleAI analysis · September 28, 2026