Conversational automation: The roadmap to intelligent customer experiences

Delivering personalized, contextual, and automated customer experiences made easy with Webex Connect's NLP, NLU & AI capabilities

Manoj Muthavarapu

Chatbots, virtual agents, NLP & AI are terms that have been used by enterprises for over a decade now. NLP and AI, like most technologies, have evolved since their inception in the early 2000s. From trying to simulate human-like interactions to now having intelligent conversations with Siri and Alexa, NLP and AI have made their way into our everyday lives.

With business-to-consumer interactions increasing rapidly, businesses must prioritize delivering slick, contextual, and personalized experiences to customers to win their business and loyalty. While AI chatbots and NLP capabilities have been primarily used to support customer service teams, there is massive potential for these technologies to be harnessed to create seamless end-to-end customer journeys.

The new normal of customer experience is personalized, interactive, media-rich communications that also extend to self-serve options. Enterprises will need to leverage advanced automation capabilities to handle contact volumes effectively. Embedding NLP and AI capabilities into customer communication strategy is the first step to managing customer inquiries quickly and effectively while bringing operational efficiency internally, saving time, and reducing costs.

AI-driven automation

A recent survey states that nearly 48% of businesses agree that implementing AI provides the opportunity to enhance customer journeys. While most companies have already adopted AI to drive better CX, many are yet to fully explore its benefits.

AI in a customer experience context has several applications ranging from customer service to enabling self-serve ensuring a quick, relevant, and personalized experience for customers. By leveraging AI to automate routine customer communications, enterprises free up their resources and reduce their operational costs by deflecting to Bot-led chats on messaging channels or self-serve FAQs portals. With close to 85% of millennials preferring to interact with businesses using live chat, conversational AI is a powerful tool for businesses looking to take their CX up a notch.

Understanding NLP and NLU

Businesses looking to offer seamless and personalized customer experiences must invest in understanding customer preferences and act accordingly.

Natural Language Processing (NLP) and Natural Language Understanding (NLU) serve as the two building blocks within AI that capture and process customer responses to turn them into actionable inputs. While NLP focuses on processing text-based inputs like identifying date and time or rectifying spelling errors in messages, NLU takes it to the next level by breaking down the intent, sentiment, and context of the message. Both capabilities are crucial for businesses that intend to engage and retain customers with relevant communications.

Conversational automation with Webex Connect

Webex Connect, our cloud-based low-code enterprise CPaaS platform enables enterprises to embed advanced NLP, NLU, and AI capabilities to automate not just customer interactions but end-to-end customer journeys across all touchpoints of a customer lifecycle. 

The NLP, NLU, and AI features together to enhance customer journeys with intelligent insights, powerful automation, and provide enterprise developers with pre-built components that can be plugged into a customer journey with a simple drag-and-drop action onto its visual flow builder.

The platform comes with two core offerings with varying levels of sophistication to enable conversational automation at scale:

  • Conversational NLP, NLU, and AI nodes

  • A bot-builder studio

Explore the bot builder capability first-hand in our Developer Sandbox.

1. Best-in-class NLP, NLU, and AI capabilities

Webex Connect offers built-in NLP, NLU, and AI nodes that can recognize and streamline customer responses with intelligent automation while breaking down customer intent and entities. The NLP nodes are most efficient for point transactions where customers respond in text (including numbers) but may contain spelling errors, junk values, inconsistent formats, or are case sensitive.

NLU nodes perform a combination of tasks where they identify and extract customer intent, mood, and sentiment based on the response and feed them back into the customer journey allowing end users or agents to have more context of the customer interactions.

A few typical functions that the NLP and NLU nodes perform are:

  • Spell check

  • Units of measurement recognition

  • Date and time recognition

  • Sentiment analysis

  • Profanity filter

The nodes when embedded within a workflow can enrich a customer journey. With their ability to detect and translate over 90 languages, these nodes make it easy for global enterprises to converse with their customers seamlessly.

NLP and NLU when used for the right use cases can help enterprises gather valuable customer behavior insights and preferences that help them to tailor their communications strategy. For instance, a UK based car finance company leverages these capabilities for a variety of use cases within its service scope. For its identification and verification (ID&V) processes, the company added NLP and NLU into its onboarding journey where the system can identify minor spelling errors or streamline various address formats to create a customer profile. It also plugged in NLP into its missed payment alerts prompting customers to respond or to chat with their agents. The customer responses received are seamlessly processed with the NLP capability and are fed back into the systems.

With Webex Connect, enterprise developers have the flexibility to choose from multiple NLP engines to suit each use case. While all the NLP & NLU functions can be used as node steps or API calls, Webex Connect's architecture allows developers to dynamically build an intelligence pipeline by assembling a series of NLP or NLU functions in a specific order and executing it in one go ensuring that as business needs evolve, developers can choose the right module based on cost, accuracy and latency tradeoffs.

2. Fully equipped chatbot builder

The out-of-the-box chatbot builder is the highest level of AI sophistication in Webex Connect. With the platform acting as a studio, the bot builder is offered as an application that can be embedded into workflows leveraging the integrations and low code visual flow builder.

Developers can build chatbots of varying complexity that can lend themselves to a variety of use cases. The bot builder supports two types of bots:

  • FAQ bots are breadth heavy bots that cover a wide range of inquiries and provide one or two mapped solutions. The interaction is transactional in nature and leverages natural language search for all its inquiries.

  • Task bots are depth heavy. These bots can hold multi-turn conversations and can perform more complex tasks like gathering data, validating customer responses, and journey fulfilment.

Both types of chatbots can leverage NLP and NLU capabilities to enhance the intelligence of customer journeys making it easier for enterprises to process customer responses. The bot builder combined with NLP and NLU becomes an out-of-the-box solution allowing enterprises to deploy these as front-of-the-house contact touchpoints to interact with customers. Taking advantage of the underlying communication channel, chatbots can be made available across various channels from SMS to more interactive rich media channels like live chat or Messenger. With its configuration-driven approach, developers can reuse the bot code logic across multiple customer journeys and channels regardless of its features.

Getting Started

Enterprises looking to embrace intelligent CPaaS to their customer communication strategy must start small and begin with identifying the right use cases and customer journeys that can be enhanced with automation. Introducing conversational automation should be a deliberate and collaborative exercise among customer-facing teams. Sales, marketing, customer service, IT, and operations teams will need to continuously work together to understand customer preferences and identify gaps within the customer lifecycle that can be enhanced with automation.

For contact centers and customer service leaders, it is important to establish a logical fallback mechanism when leading with AI-powered chatbots for customer service. While chatbots can provide great front-of-the-house customer service, it is crucial to offer the option to customers to deflect to an agent or another channel of their choice. Enterprise IT needs to continually evaluate and feedback into the bot to enhance its performance and widen the scope of inquiries it can address, enabling contact center agents to focus on more complex and urgent inquiries.

By deploying chatbots as the first line of contact, enterprises can aim to provide 24/7 support, intimate customers of their working and seamlessly hand over complex or urgent cases to human agents. The AI-powered bots can continue to grow and enhance in-life using regression testing and Human-in-the-loop (HITL) learning where human feedback can be incorporated to enhance the scope of the chatbots’ intelligence. Developers can update their knowledge base, add relevant tags or articles, and feed their performance back into the algorithm that allows the chatbot to offer smarter, relevant, and more contextual interactions needing lesser manual intervention.


Webex Connect is uniquely equipped to cater to various parts of the business, like contact centers, customer experience professionals and IT teams enabling them to build and deploy customer journeys of varying complexity with its embedded NLP, NLU, and AI capabilities. Enterprise developers and users can significantly accelerate the delivery of rich, intelligent, and interactive customer experiences. Its best-in-class, out-of-the-box NLP and AI offerings empower enterprises to deliver smarter customer interactions faster.

Want to know more about Webex Connect? Try the platform now using our developer sandbox and explore all our NLP, NLU, and AI offerings.