The Next Frontier of AI: Enabling AI Agents to Get Work Done Across the Enterprise

Over the past several years, the industry has made extraordinary progress in teaching artificial intelligence (AI) to understand language.

AI is now proficient in identifying intent, summarizing conversations, analyzing sentiment and generating increasingly sophisticated and humanlike responses. As a result, organizations are beginning to rethink how AI can be applied across customer experiences and business processes.

Customers contact businesses because they want something done. They want a billing issue resolved, an order updated, a claim processed, a shipment expedited or an account restored.

The next phase of AI will be defined by its ability to move beyond understanding customers to orchestrating the work and resources (both human and AI) needed to drive the outcomes they’re looking for.

This is the shift from understanding and engaging with customers to actioning and governing work across the enterprise.

The industry is moving quickly in this direction. Gartner^® predicts that “by 2028, more than 60% of enterprise customer service interactions will be managed end-to-end by agentic AI, up from 20% in 2026”.¹

The question is whether organizations can build the ecosystem connectivity, workflow automation, observability and governance required to operationalize it responsibly.

Turning Intent into Action that Drives Outcomes

Understanding the customer’s intent is an essential first step, but it isn’t the goal. Delivering the customer’s outcome often requires several actions that involve coordinating information, workflows and decisions across multiple systems and business functions.

That distinction becomes increasingly important as organizations adopt agentic AI.

Customer intent to outcome

Customer says

“My package has not arrived.”

Customer intent

Resolve a delivery issue

Back office workflows

• Validate order & delivery status
• Investigate with carrier

• Update shipment/reschedule
• Notify customer

Enterprise systems

• CRM — Customer profile, order history
• OMS — Order status, shipment tracking

• ERP — Inventory, warehouse, fulfillment
• WMS — Warehouse management system

Outcome delivered

Delivery issue resolved

Customer is informed and shipment is rescheduled.

Business impact: Higher customer satisfaction and loyalty

A customer whose package is delayed doesn’t want an apology or to even understand the cause. They just want their package. A customer who has been charged twice doesn’t want a summary of their account history. They want a refund on the extra charge.

Many organizations still struggle to move from understanding what a customer needs to actually resolving it. Customer journeys rarely exist within a single system, and the work behind a request often depends on data, workflows and approvals spread across the enterprise. As organizations implement agentic AI, systems must be securely connected so they can access information, initiate processes and coordinate work to fulfill requests.

Resolving a customer issue often requires orchestrating actions across CRM platforms, billing systems, order management applications, ERP systems and countless other business tools across many different corporate departments. These systems frequently operate in silos, with disconnected workflows, manual processes and layers of operational complexity between customer engagement and execution.

This challenge has become one of the biggest barriers to realizing the promise of agentic AI.

Building the Foundation for Autonomous Experiences

Enabling organizations to turn intent into outcomes is one of the reasons we recently acquired Pinkfish, a leading agentic workflow orchestration company.

We wanted to help enterprises address one of the biggest obstacles to agentic orchestration: fragmented technology environments where the systems, data and workflows required to drive outcomes remain disconnected and difficult to coordinate.

As AI agents become more capable, enterprise connectivity is becoming a strategic requirement, rather than a technical implementation detail. AI agents need a consistent way to discover tools, access information and take action across enterprise systems while maintaining governance and control.

Pinkfish brings broad enterprise connectivity and workflow automation capabilities to the Genesys Cloud™ platform, including access to more than 500 integrations and 25,000 MCP tools across CRM, ERP, IT, HR, order management, billing and other business applications. These capabilities will expand the MCP ecosystem within Genesys Cloud.

MCP is becoming increasingly important because it provides a standardized way for AI agents to discover, access and use enterprise tools. Rather than relying on custom integrations for every application and workflow, organizations can expose business capabilities through a common interface that AI agents can securely access. This simplifies integration and improves interoperability, governance and scalability as organizations deploy more agents.

For organizations pursuing agentic orchestration, that matters.

For example, when a customer contacts a retailer about a delayed order, an AI agent should be able to do more than explain the problem. It should be able to verify the order, evaluate shipping options, understand the customer account, apply an approved service credit, upgrade delivery and notify the customer while operating within defined business guardrails. That is the type of outcome-driven experience organizations are increasingly looking to deliver.

Just as importantly, natural language-driven workflow automation lowers the barrier to creating sophisticated AI-powered workflows. Instead of relying on scarce technical resources and project prioritization to build and maintain integrations, organizations and business users can rapidly create, adapt and manage workflows as business requirements evolve.

The Future of AI Will Be Defined by Enterprise Execution

Over the past several years, the industry has made extraordinary progress in making AI more intelligent. The next challenge is putting AI to work enterprise-wide.

Solving that challenge will determine how quickly organizations move from experimenting with agentic AI to fully operationalizing it. For organizations, that means enabling AI to securely access information, coordinate actions and complete work across the systems that power their business. That is what will determine whether agentic AI remains an interesting technology or becomes a transformational one.

At Genesys, we’re helping organizations build the foundation for agentic orchestration. For our customers, that foundation will be critical to turning the promise of agentic AI into meaningful business outcomes and advancing the future of autonomous customer experiences.

Availability 

Genesys expects Pinkfish capabilities to become available to Genesys Cloud customers through the Genesys AppFoundry^® Marketplace by the end of the second quarter of the company’s fiscal year 2027 (May 1-July 31, 2026). The company expects to begin introducing Pinkfish capabilities natively within Genesys Cloud by the end of this fiscal year (Feb 1, 2026-Jan. 31, 2027). 

¹ Source: Gartner, “Emerging Tech: Agentic AI Adoption Trends and Future Industry Opportunities”, Christine Tenneson, Ehtan Cai, Kiumarse Zamanian, Aakanksha Bansal, Alfredo Ramirez IV, 11 June 2026  

GARTNER is a trademark of Gartner, Inc. and/or its affiliates.