The conversation around AI and customer service has centered on one thing: reducing costs. Executives have prioritized efficiency gains over customer experiences. They’ve aimed AI investments at automating service and scaling efficiency and cost-cutting, all with the semblance of being transformative. But that’s a finite strategy. Winning companies are going to make service intelligent, personal, and experiential.
For decades, CRM was the repository of customer truth. It was valuable, but siloed. Workflow automation changed that. Suddenly, the system was orchestrating interactions instead of just recording them. Routing decisions, initiating workflows, and triggering actions across fulfillment, billing, inventory, HR, IT, and beyond. With AI, those workflows are now intelligent. They can learn, decide, and understand context in ways that were never possible before.
In an era of AI and agentic AI, experiences no longer have to be disparate, disconnected, or siloed. Experiences can now be connected, personalized, and optimized all in the one customer journey, north to south, and east to west, across the enterprise. That’s an operating model transformation that c-suite leaders need to address today.
In Infinite: How Visionary Leaders Transform Today’s Businesses into AI-Forward Companies, we describe this as the shift from org charts to work charts. The org chart tells you who reports to whom. The work chart tells you how value flows, and what human roles, agent roles, decisions, and outcomes matter.
For customer experience, the work chart means that CRM sits in the middle of sales, service, fulfillment, and back-office work. AI agents become the interface. But the architecture that wins combines deterministic governance (clear rules, policies, guardrails) with probabilistic reasoning. CRM transforms from a system of record into an orchestrated system of action. The customer, as a result, has a better overall experience.
The Platform Problem Nobody’s Talking About
Most companies run their business on cloud platforms, collaboration tools, custom applications, and specialized point solutions. CRM is one node in a much larger network.
But there’s a fundamental problem with this siloed approach. If each vendor introduces its own agent with its own policies and logic, and no unified platform, the result is agent sprawl. Companies risk overlapping identities, inconsistent controls, duplicate context, conflicting actions, and fragmented audit trails.
Think of it this way. A customer calls with a billing question. An agent in your CRM knows their history. But it doesn’t know that an agent in your fulfillment system just took an action that renders the CRM’s recommendation irrelevant. Security policies are disconnected. When something goes wrong, you have no clear lineage of decisions. You can’t explain why the customer was treated a certain way. You can’t hold anyone, human or AI, accountable.
