Cisco Launches Dialog Agentic Framework for Webex Enterprise Platform

Cisco is updating its Webex cloud collaboration platform to better accommodate AI agents, introducing a new toolset called Dialog. The company describes Dialog as an "agentic harness" that directs AI to handle customer interactions, promising to turn fragmented interactions into continuous customer relationships.

Dialog Keeps Agents Working After Conversations End

According to Cisco, Dialog tasks various agents with keeping the consumer happy, as they will keep working on the customer's behalf after a conversation ends. Agents under Dialog's umbrella can coordinate across people, agents and backend systems. The company also says the platform continuously refines agent performance with every interaction.

AI Agents Join Spaces, Meetings and Calls

The updated Webex will let anyone in a workspace directly invite AI agents into spaces, meetings and calls to execute complex, multi-step work across the platform and in third-party applications. Cisco gives an example of an agent quickly pulling up statistics during a presentation and updating a PowerPoint on the fly.

The platform also speeds up the onboarding process by treating AI agents like new teammates. Agents are given the opportunity to peruse existing employee handbooks, knowledge bases and operating procedures. Cisco promises a safety-forward experience via Splunk, a data analysis platform it acquired a few years ago, though the company notes this is a new feature and issues could still emerge.

Agent Reliability Remains an Industry Challenge

Cisco's announcement comes amid ongoing concerns about AI agent reliability. Researchers at Carnegie Mellon University found earlier this year that even the best-performing AI agent at the time, Google's Gemini 2.5 Pro, failed to complete real-world office tasks 70 percent of the time.

Multiple studies have indicated that AI agents are still prone to mistakes which lead to cascading failures. The failure rate sits somewhere between 60 and 70 percent with regard to multi-step tasks, because a single mistake during one step can throw the whole process off. Even an agent with a 95 percent per-step reliability rate will only succeed around 36 percent of the time when dealing with a 20-step workflow. Even a 99 percent per-step reliability rate seems risky when money and a company's reputation is on the line, though performance will likely improve over time.

Sources: Engadget