The request arrives
A customer calls or messages, and MCM AI reads why from their first few words. The reason, the language and the account behind the contact are known before anyone is chosen.
MCM AI feature
The next call goes to the agent likeliest to resolve it.
MCM AI pairs each customer with the agent most likely to resolve their request. It weighs why they are calling, who they are and who is free right now, then makes the match before the phone rings.
Illustrative example
How it decides
Six signals behind every match.
A good match is more than a free line. Pick a signal to see what it adds.
Why they are calling
MCM AI reads the reason from the customer’s own first words, so the match starts from what they actually need.
Signal 1 of 6 · Learn more
Side by side
Skills set the pool. Prediction picks from it.
The two work together. Skills-based routing decides who is eligible, and predictive routing chooses the best of them.
Skills-based routing
Predictive routing
How the agent is chosen
Skills-based:The request goes to agents tagged with the right skills and languages.
Predictive:Each customer is paired with the agent likeliest to resolve their request.
Who does the thinking
Skills-based:You define the skills and the rules in advance.
Predictive:MCM AI weighs the signals for every contact, inside the rules you set.
When queues get busy
Skills-based:The rules you wrote keep applying as they are.
Predictive:MCM AI adjusts in real time to keep waits short.
Best suited to
Skills-based:Clear, stable groups of skills.
Predictive:Mixed queues where the best fit changes from call to call.
A call, matched
Follow one request from ring to rep.
The field is narrowed
Agents who are away, already full or missing the right language drop out. What remains is the group of people who can genuinely help this customer right now.
The likeliest resolver is picked
From that group, MCM AI chooses the agent most likely to resolve this kind of request for this kind of customer, using what it knows about both sides of the match.
The agent picks up informed
The reason for the call and the customer’s details are already on screen, so the first words your agent says are about the solution, not the introduction.
Illustrative example
Alongside the rest
Part of a routing setup you control.
Predictive routing slots into the same visual flow builder as the rest of your call handling.
01
Priority customers, still well matched
VIP and urgent customers go to the front of the line, and predictive routing still looks for the best agent for them once they get there.
02
Overflow and callbacks as a safety net
When every suitable agent is busy, overflow rules or a queue callback keep customers from waiting on the line, and the match is made when someone frees up.
03
Your CRM in the decision
Customer details from your CRM feed the match, so routing can greet people by name and take account of who they are before it picks an agent.
Pair it with queue callback, and design the whole flow in the Smart IVR & Routing builder, or build agents in the no-code AI builder.
Outcomes
Every customer meets the right person.
Right agent first
Customers reach the person best placed to help, not the next one in line.
Shorter waits
Live queue load shapes every match, so work spreads sensibly across your team.
Fewer hand-offs
When the first agent is the right one, customers are not passed around to find them.
Part of
Smart IVR & Routing
The product this feature belongs to. Related: AI Voice & Chat Agents and Omnichannel Queue.