Cisco conversational AI powered by Amazon Lex

By: Aaron Keeton, Business Development Manager – Cisco
By: Jessica Smith, Director Product Marketing – Cisco
By: Soundarya Muthuvel, Product Manager Webex Developer Experience – Cisco
By: Ravi Thakur, Sr. Solutions Architect – AWS
By: Satish Subramanian, Manager Partner Solutions Architects – AWS

When customers call your support line, they’re not thinking about the technology behind it. They want answers – fast, accurate, and without the runaround.

But contact center leaders face a familiar challenge: you’ve invested in Cisco Webex Contact Center, your teams have expertise around it, and it delivers the experience your customers want. Meanwhile, conversational AI is reshaping what customers expect-instant responses, natural conversations, and uninterrupted availability without waiting in line. The pressure to modernize is real but ripping out what’s working isn’t a beneficial strategy.

That’s why Cisco and Amazon Web Services (AWS) took a different approach. They brought AWS AI capabilities into the Cisco environment, so you can modernize without starting over.

The new open source Amazon Lex connector for Webex Contact Center brings intelligent virtual agents into your existing Cisco environment, without rebuilding from scratch. Amazon Lex now plugs directly into Cisco’s Bring Your Own Virtual Agent (BYOVA) framework. With the Amazon Lex integration, you can access a powerful AI-driven virtual agent alongside your contact center workflows without any additional work, providing consistency across voice experiences. Your customers get natural, voice-driven interactions. Your team retains their existing Webex workflows, reporting dashboards, and agent desktop configurations while gaining AI capabilities that would otherwise require months of integration work.

For the more than 10,000 organizations running Cisco Webex Contact Center and Contact Center Enterprise today, this solution removes the trade-off between protecting your investment and adopting AI. You can deploy across many AWS Regions globally, and it supports over 25 languages and automates routine inquiries so that your agents can focus on conversations that really need a human.

Here’s how the integration works, where it fits across Cisco Webex Contact Center and Cisco Contact Center Enterprise, and where customers are already seeing results.

Cisco and AWS integration architecture

This integration connects your Cisco Webex Contact Center to AWS AI services through an open-source connector. The architecture comprises three integrated layers:

• Customer interaction layer – Webex Contact Center and Contact Center Enterprise connect through Webex Contact Center AI (CCAI), which manages the virtual agent integration and handles escalations routing to an agent when human assistance is needed.
• Connector layer – The open-source connector serves as the technical bridge between platforms, authenticating between Webex and AWS environments, managing real-time gRPC media streams for audio, and orchestrating the conversation flow between platforms.
• Conversational AI layer – Amazon Lex provides conversational intelligence through automatic speech recognition (ASR), natural language understanding (NLU), and integration with large language models (LLMs) using Amazon Bedrock. Speech synthesis is handled by Amazon Polly, delivering natural-sounding responses in dozens of voices across multiple languages.

This architecture is illustrated in the following diagram.

Figure 1: Cisco Webex Contact Center integration with AWS AI services

Customer call flow

When a customer calls in to the Webex Contact Center, the following flow takes place:

1. Audio streams in – The caller’s voice flows from Webex Contact Center through Webex CCAI to the open-source connector. The connector establishes a gRPC media stream to carry the audio in real-time to AWS services. This stream also enables bidirectional metadata exchange, including caller information, historical journey, and subscription data between Webex CC and AWS.
2. Amazon Lex interprets intent – Amazon Lex processes the customer’s speech through three stages:
   1. Speech recognition – ASR converts voice to text, supporting more than 25+ languages
   2. Intent understanding – NLU determines what the customer wants, now enhanced with LLM-powered assisted NLU for faster, description-based intent classification
   3. Response generation – For straightforward requests, Amazon Lex responds from configured intents. For complex queries requiring nuanced responses, Amazon Lex invokes Amazon Bedrock foundation models (FMs) to generate contextual, conversational replies that go beyond scripted answers.
3. Response flows back – Amazon Lex formulates the response and passes it to Amazon Polly. Amazon Polly transforms text into natural-sounding speech using neural text-to-speech (NTTS) technology, with support for Speech Synthesis Markup Language (SSML) to control pronunciation, emphasis, and speaking rate. The audio streams back through the connector to the customer using Webex Contact Center.
4. Escalation when needed – If the virtual agent can’t resolve the issue, the call routes back through Webex CCAI to a live agent within your Cisco Webex Contact Center environment. The agent receives full conversation context—including customer intent, information gathered, and dialog history—eliminating the need for customers to repeat themselves.

This architecture supports multi-turn conversations with contextual awareness requests like “I want to change my appointment from Tuesday to Thursday,” which work without requiring customers to start over.

The customer call flow is illustrated in the following flow diagram.

Figure 2: Cisco Webex Contact Center and Amazon Lex integration call flow

This solution integrates with AWS in the following ways:

• AI/ML services – Amazon Lex handles ASR and NLU, converting customer speech to text and determining intent. Amazon Polly delivers lifelike voice responses using Neural Text-to-Speech (NTTS).
• Agentic developer tools – Amazon Bedrock AgentCore provides the framework for building and orchestrating AI agents.
• Generative AI with Amazon Bedrock – For more complex interactions, Amazon Bedrock connects your virtual agent to FMs, enabling responses that go beyond scripted intents.

The value of this architecture lies in the control it gives you. The Amazon Lex chat assistant lives in your AWS account. You control the intents, responses, and data. The solution is designed for responsive, natural conversations. Security is through cross-account authentication, which means your customer data stays in your environment. This solution is available in AWS Regions that support Amazon Lex endpoints to meet your compliance and data residency requirements.

Industry specific use cases and key benefits

Organizations across industries can find benefits in pursuing this approach, including in healthcare, finance, and government.

Healthcare organizations can enhance patient experience with intelligent routing and reduced wait times by consolidating outdated systems that need a live agent for every task. The solution is integrated with scheduling systems for seamless appointment management using automated booking, rescheduling, or canceling.

Finance organizations can improve customer self-service in situations where they don’t need to talk to a live agent by using intelligent interactive voice response (IVR), chat-based assistants, and virtual agents. Customers can check balances, update personal details, or reset passwords without a live agent. Reduce complexity through unified customer experience platforms and cloud-based solutions.

Governments can operate in Cisco’s secure, FedRAMP-authorized data centers for reliability and compliance. This solution enables migration from on premises to the cloud at the agency’s pace with an intuitive administration portal for faster operations and reduced IT burden.

Business outcomes

Conversational AI transforms routine customer interactions, including billing inquiries, order tracking, and appointment scheduling into seamless self-service experiences. Customers get immediate responses without waiting in queue, while your contact center achieves higher first-contact resolution rates. The result is consistent, quality support available around the clock without proportional staffing costs.

Similar efficiency gains apply to order status and appointment scheduling interactions.

Natural language understanding in Amazon Lex goes beyond recognizing what customers say, it understands the intent of the customers. By connecting to customer data, the virtual agent delivers contextual responses. It recognizes repeat callers, references recent interactions, and anticipates needs. This intelligent, empathetic approach builds loyalty by making every customer feel known, not processed.

Conclusion

In today’s Cisco Webex Contact Center deployments, organizations face a defining moment. The need to address how to safeguard existing investments and resources but adopt AI capabilities to get natural, voice-driven interactions. Through the open-source Amazon Lex connector for Webex Contact Center, we’ve demonstrated how customers can bring powerful AI-driven intelligent virtual agents into their existing Cisco environment without significant investments or re-architecture.

Organizations implementing this open-source Amazon Lex connector show improved AI agent efficiency and enhanced customer experience. This solution integrates with proven AWS AI capabilities to improve every interaction and maintain context and intent across complete conversations, improving your customer experience.

Refer to this blog for step by step instructions on how to integrate Webex Contact Center with open-source Amazon Lex connector.

Ready to transform your customer experience strategy? Connect with Cisco and your AWS account team today to schedule a personalized consultation and discover how this integration drives customer satisfaction and business growth for your organization.

.

Cisco Systems – AWS Partner Spotlight

Cisco Systems is an AWS Specialization Partner that helps customers optimize their cloud strategy by bringing together networking, security, analytics and management.

Contact Cisco | Partner Overview | AWS Marketplace