Cisco Starts to Embed AI Agents Into Its Webex Contact Center

Cisco has started to add native AI agents to its Webex Contact Center.

AI agents function within systems, reasoning on data, adapting, and actioning tasks on behalf of users. 

Unfortunately, in the contact center space, some still consider an AI agent to be another word for a chatbot. That’s not the case. 

Yes, an AI agent can interact with customers and try to resolve their queries. Nevertheless, the use cases expand much further.

Recognizing this, Cisco plans to embed additional agents into its CCaaS solution, alongside its customer-facing Webex AI Agent. 

The rollout of these agents started late last year when Cisco expanded its Webex AI Quality Management solution, introducing an AI agent that monitors and coaches the customer-facing AI agent.

Sharing an example of this in action, Snorre Kjesbu, SVP and GM of Collaboration at Cisco, told ZK Research: “In financial services, a person calls in, talks to an AI agent, and that AI agent did not pitch the new credit card at three minutes and 27 seconds (i.e., the right moment) in the call. [Now], you have another AI agent also monitoring that.”

Cisco didn’t formally market this as an AI agent. Instead, “actionable recommendations and performance optimization for AI agents” came part and parcel with the new-look Webex AI Quality Management solution. Nevertheless, it exemplifies how Cisco is starting to embed AI agents across the Webex Contact Center.

In 2026, Kjesbu plans to go deeper, but has pledged to keep a focus on meeting explicit outcomes."  

The key thing in 2026 is to go from technology excitement - and excitement about a lot of these agents - to actually tangible outcomes.

Snorre Kjesbu

SVP and GM of Collaboration at Cisco

Indeed, there are several areas where Cisco and its CCaaS competitors could deploy AI agents within their software, and they may consider building AI agents like those shared below.

The Possibilities of AI Agents within Contact Center Software 

There are many ways in which AI agents could augment the workflows of live service reps, whether that’s by intelligently retrieving information from adjacent applications, auto-filling forms, or completing tasks like sending personalized email confirmations. Over time, these use cases will become increasingly sophisticated. 

One particularly fascinating possibility is monitoring signals across live rep interactions and spotting employees who have endured several complex calls on the spin. An AI agent could then proactively offer the rep a reset break to avoid burnout.

Cisco already does this with software. But, in the future, an AI agent also opens up new possibilities, such as routing reps a contact that’s likely to be simpler, per the customer’s stated intent, in place of that break. This could immediately help live reps get their mojo back.

AI agents could also augment the triaging process by tracking contact center workforce management (WFM) data to identify when an employee’s shift is about to end and routing an easier call to them, safeguarding well-being and morale.

Another possible WFM use case is an AI agent that monitors contact volumes, service level, and other operational metrics and autonomously performs key intraday management tasks by following standard operating procedures (SOPs).  

Reporting is also an area AI agents could transform, allowing supervisors and managers to spin up dashboards with written prompts and interact with the numbers they see. 

Some of these use cases may appear a distant prospect. Yet, as Cisco and its competitors embed AI agents into their platforms and limit the need for custom development, the technology becomes much more tangible.

Supporting Human Contact Center Reps in a World of AI Agents

As AI increasingly automates transactional contacts, the human role becomes more complex. After all, the simple calls they’d use as a breather start to fall away. Instead, it’s tough call after tough call.

AI assistants and agents can help by suggesting next actions - and offering to perform them - when the rep is unsure of the best route forward. 

Yet, many live reps sometimes avoid leveraging assistive AI solutions, as they often do when new technologies come into the contact center.

Against this backdrop, service leaders should consider slowly shifting away from judging human agents only on speed and volume to also measuring how well they work with AI.

That may mean tracking new metrics, like whether reps appropriately accept or challenge AI suggestions, rather than blindly following or ignoring them. 

Additionally, leaders may wish to consider new metrics that dig deeper into the skills of human reps. 

For instance, some contact centers have started attaching complexity scores to tickets and monitoring the reps’ corresponding quality scores to monitor rep capability.

However, such a metric can be somewhat arbitrary. As such, tracking training time and skills coverage is perhaps the better option.

Nevertheless, as CCaaS vendors expand their assistive AI portfolios, proactively offering such guidance so contact centers can maximize the value of their deployments is something they should consider closely.

After all, as the AI curve accelerates, providers can differentiate by being more consultative and holding customers' hands closely through the transition.