Customer support · Production

ElevenLabs powers voice technology for Cisco's Webex AI Agent

The problem

Enterprise customer service falls far short of consumer expectations: only 25% of customers are very satisfied with their last service engagement, 94% have abandoned interactions due to poor experiences, and traditional chatbots have frustrated customers with rigid experiences and an inability to interpret natural language.

First attempt

Traditional chatbots frustrated customers with rigid experiences and an inability to interpret natural language.

Workflow diagram · grounded in source
1
Customer initiates support
trigger
“bringing voice-first, agentic customer support to the enterprise”
2
LLM understands customer needs
ai_action
“Webex AI Agent uses large language models (LLMs) to deliver more human-like customer experiences that understand an individual's needs, remember their history, and adapt to their preferences”
3
Voice AI adapts emotional delivery
ai_action
“voice AI can respond to a customer's emotional cues and adapt its delivery — sounding warm and welcoming when needed, or serious and empathetic when the situation calls for it”
4
Backend system integration
integration
“enabling seamless integration with custom knowledge bases and critical backend systems such as CRM, ERP, and HR platforms”
5
Human-sounding AI voice response
output
“deliver voice interactions that sound and feel like conversing with a real person—thanks to the agent's human-like intonation, inflections and rhythm”
Reported outcome

The Webex AI Agent, powered by ElevenLabs voice technology and LLMs, delivers human-like voice interactions that understand customer needs, remember history, adapt to emotional cues, and integrate with enterprise backend systems at enterprise-grade scale and security.

Reported metrics
Customers very satisfied with last customer service engagement25%
Customers who abandoned interactions due to poor experiences94%
Customers who would switch brands due to sub-par customer service95%
Customers frustrated by having to repeatedly explain their issue54%
Show all 6 reported metrics
customers very satisfied with last customer service engagement25%
customers who abandoned interactions due to poor experiences94%
customers who would switch brands due to sub-par customer service95%
customers frustrated by having to repeatedly explain their issue54%
customers frustrated by voice or chatbots not intelligent enough to help48%
customers frustrated by being kept on hold too long31%
Reported stack
ElevenLabsWebex AI Agentlarge language models (LLMs)Text to SpeechCRMERPHR platformscustom knowledge bases
Source
https://elevenlabs.io/blog/cisco-webex
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

The Webex AI Agent, powered by ElevenLabs voice technology and LLMs, delivers human-like voice interactions that understand customer needs, remember history, adapt to emotional cues, and integrate with enterprise back…

What tools did this team use?

ElevenLabs, Webex AI Agent, large language models (LLMs), Text to Speech, CRM, ERP, HR platforms, custom knowledge bases.

What results were reported?

Customers very satisfied with last customer service engagement: 25%; Customers who abandoned interactions due to poor experiences: 94%; Customers who would switch brands due to sub-par customer service: 95%; Customers frustrated by having to repeatedly explain their issue: 54% (source-reported, not independently verified).

What failed first in this deployment?

Traditional chatbots frustrated customers with rigid experiences and an inability to interpret natural language.

How is this customer support AI workflow structured?

Customer initiates support → LLM understands customer needs → Voice AI adapts emotional delivery → Backend system integration → Human-sounding AI voice response.

Related customer support cases