At a glance
Industry
Communications
Location
North America
Challenge
Create an innovative conversational AI channel for Cox Communications customer support.
Success highlights
- Migration to Google Cloud Contact Center as a Service (CCaaS)
- Delivery of gen AI-powered chatbot in three months
- ~46% resolution rate for Oliver chatbot
- ~54% resolution rate for conversational IVR
- 17% improvement in first-contact resolution for chatbot
- 8x increase in customer use of the digital channel
The challenge
Each month, Cox Communications’ customer support manages upward of 700,000 interactions with the company’s residential and business customers. Conversational AI channels are a key initiative for Cox, a leading US telecommunications provider. To make those channels scalable and more useful for customers, the company sought to move its chatbot ecosystem to an AI-first platform that’s adaptive and easier to optimize in real time.
In high-volume telecom environments like Cox’s, every customer interaction counts, especially with chatbots. Bot performance directly impacts customer retention—and agentic and generative AI raise the bar for conversational bots, with customers expecting accurate, relevant responses.
Our approach
Together, Cognizant and Cox built a scalable, AI-first contact center. The project represents the first customer-facing AI implementation within Cox.
We kicked off the project by migrating Oliver, Cox’s virtual assistant, to Google CCaaS. Key to this effort was replacing the company’s existing customer service platform, Khoros, with UJET, an AI-powered cloud contact center platform that enables agent-assisted support.
Key components of the Oliver chatbot solution included:
- Hybrid AI approach with Google Dialogflow CX and playbooks, using rule-based workflows for structured customer requests and gen AI for more complex conversational interactions
- Intelligent routing to route customer inquiries to rule-based workflows or generative AI responses based on the nature and complexity of the request
- SmartHelp self-service troubleshooting that enabled customers to resolve common technical issues without agent intervention
- Bill pay for secure online payment processing and automated eligibility checks for flexible payment plans
Gen AI-powered IVR
Development of the AI-first contact center included establishing a unified AI model—a key step in supporting Cox’s goal of rapid business value. That common AI foundation enabled Cox to launch Olive, a gen-AI-powered voice bot for its IVR system built on the same platform, just three months after Oliver's re-launch. The voice bot debuted with a single intent: allowing customers to schedule appointments for service. Its rapid adoption led Cox to quickly add several dozen more intents to the new IVR system.
Business outcomes
Cox relaunched the Oliver chatbot three months after signing the contract with Cognizant. Thanks to the hybrid model’s balance of deterministic and agentic AI capabilities, adoption among customers was high—and the business impact for Cox was immediate. Within a week, first-contact resolution improved 10%, and within a year, it jumped to 17%. The bot’s increased context awareness drove an 8x increase in customers’ use of the chatbot channel. Equally important, the bot’s continuous feedback loop enabled Cox to implement more than 500 improvements.
