Most contact center platforms still treat routing as the defining moment in a conversation. But the customer journey continues to evolve after that first decision. Intent changes. Wait times increase. Workforce availability shifts. Business priorities move. Each of these real-time signals shapes the experience. This creates an opportunity to manage the conversation lifecycle more effectively — not as a one-time routing event, but as a dynamic journey guided by live context from both the customer and the contact center.
Contact centers are constantly balancing customer need with workforce capacity. ACD has always helped manage that balance, but when conditions change, service levels, satisfaction, and trust change with them. That balance cannot be set once and left alone. It has to be sensed and adjusted in real time. Otherwise, supervisors are forced to step in manually, often after the moment to act has passed. What contact centers need is not more static logic, but a system that can continuously sense, decide, and respond as the conversation unfolds.
Introducing conversation orchestration in Dynamics 365 Contact Center
In April we introduced conversation orchestration, a new tool within Service Operations Agent. It extends traditional routing by enabling the contact center to respond dynamically as conditions change. It allows administrators to define business policies in natural language using a model of events, conditions, and actions.
Conversation orchestration continuously monitors signals across the conversation lifecycle, evaluates them against the configured playbook, and triggers the appropriate response. The result is a more adaptive operating model that helps contact centers manage change in real time and deliver more consistent customer outcomes. We will continue to expand conversation orchestration over time by adding capabilities — each one targeting a specific, high-impact contact center use case. The first two are available today.
Policies written in natural language. Executed at scale.
The hardest part of building intelligent contact center behavior has never been the technology. It has been the configuration. Rule trees that take weeks to build, require IT to modify, and break in ways that are invisible until a customer notices. The expertise of your best operations leader, locked inside a configuration file that nobody else can read or change.
Conversation orchestration replaces that with playbooks — natural-language instructions that define what should happen at each stage of the conversation journey. An admin does not configure a rule. They write an intention:
“If a premium customer is waiting in the queue and no support reps are available, increase their priority over time.”

“If a premium customer is waiting in the queue unassigned, immediately offer a callback.”

That is the entire instruction. Conversation orchestration handles the rest: reading the conversation context in real time, evaluating the conditions against customer attributes from CRM, executing the action, and doing it with consistency that no manual process can match.
Guided templates make common scenarios fast to configure. Built-in validation catches conflicts before anything goes live. What used to require a specialist and a change management process can now be authored, validated, and published by an operations admin in minutes.
This is the difference between a system that routes and a system that orchestrates.
The first set of capabilities: available now in Public Preview
We are launching conversation orchestration with first set of capabilities in public preview. They are not arbitrary starting points. They address the two most universal failures in contact center queue management — the ones that affect nearly every enterprise contact center, every single day.
Dynamic prioritization
Priority set at queue entry and never revisited is one of the most quietly damaging problems in contact center operations. A customer waiting ten minutes is competing for the same representative as someone who just arrived. A conversation transferred mid-journey — carrying all the context of what came before it — resets to the destination queue’s default priority as if it had just called in.
Dynamic prioritization fixes this by keeping priority evaluation alive throughout the conversation. As wait time accumulates, as transfers occur, as customer tier is confirmed from CRM — the priority score adjusts automatically. High-value customers and long-waiting callers move forward without anyone monitoring the queue manually. Conversation orchestration does what the best supervisor would do, but for every conversation simultaneously.
Overflow based on customer service availability
When no customer service representatives are available, the standard contact center response is silence. The customer waits. Maybe a representative comes online. Maybe they do not. Overflow rules exist in some systems, but they wait for a fixed timer to expire before doing anything — which means every customer experiences the same minimum wait regardless of whether help is actually coming.
Overflow based on customer service representative availability acts the moment it is needed. As soon as a conversation enters a queue with no eligible representatives — accounting for skills, capacity, and assignment logic — the overflow playbook fires. And the action it takes is not the same for every customer. Context variables determine the outcome: one segment transfers to a backup team, another receives a callback offer, a third gets a graceful closing message. Every customer gets a response. None of them wait in silence.
Try It Today
These two capabilities are the first step in a broader orchestration platform. We will be enabling more scenarios enabling you to automate the contact center behavior.
Conversation orchestration is available now in public preview for voice and live chat channels in Dynamics 365 Contact Center. Access it through Copilot Service admin center under Get started > Contact Center Agents > Service Operations Agent(preview) > conversation orchestration (Preview).
Start with a template, configure your first playbook, and see what it means for a contact center to stay with every conversation — not just route it.




