Workforce Orchestration is becoming a priority for CX leaders as AI agents move from isolated pilots into live customer operations.
For Gal Rimon, CEO and Founder at Centrical, the issue is that AI agents are entering customer operations through different systems, teams, and workflows without one clear model for ownership, measurement, collaboration, and frontline execution.
AI agents are already supporting human agents with admin work, transcriptions, summaries, notes, and actions. They are also helping customers complete simple tasks, such as changing bookings or handling routine service requests, while other AI agents are beginning to operate inside larger back-office processes. Rimon described the Agent Zoo as a problem of disconnected optimization:
“The challenge is that each AI agent is trying to optimize themselves instead of the customer experience or the employee experience, and this is what I call the agent zoo.”
For CX leaders, that raises a difficult question of ownership. If different AI agents sit inside different platforms, teams, and workflows, who is responsible for the performance of the overall customer journey?
Workforce Orchestration Can Bring Order to the Agent Zoo
The Agent Zoo describes a CX environment where multiple AI agents operate across different touchpoints without one shared performance model.
Each AI agent may perform well inside its own narrow task, but still create a poor end-to-end experience for the customer or employee. Rimon pointed to the risk of AI agents optimizing locally while the wider journey breaks down across handoffs, escalation points, and back-end processes.
He described a recent airline experience where an AI agent understood his request and handled the conversation well, but then sent him back into the same broken application flow. The AI worked in isolation, but the journey still failed.
Rimon framed that as an operational risk around handoffs, not a failure of AI capability. CX leaders may see strong proof-of-concept results, but production environments expose whether AI agents can support the full customer journey.
Fragmented Systems Widen the Execution Gap
Rimon also connected AI agent sprawl to a wider problem: the distance between strategy and execution.
Business leaders may define a goal, create a plan, and identify the desired customer outcome. The challenge starts when that goal has to reach multiple channels, human agents, AI agents, supervisors, applications, and processes. Rimon argued that visibility alone does not solve the execution problem:
“The main thing is not where the problems are, but actually how to fix the problems, how to bridge the strategy to the behaviors across humans and AI to achieve the intended outcomes."
Centrical calls this the behavior gap. In Rimon’s view, the central challenge is how to bridge from strategy to the right behaviors across human and AI agents.
That distinction is important for CX leaders. More AI does not automatically create better service. Better service comes when AI agents, employees, and managers operate around the same outcomes, guardrails, and feedback loops.
