/
5 mins read

Delivering Immersive CX with ChatGPT: Cristina Fonseca, Head of AI, Zendesk 

1. What is the role of generative AI like ChatGPT in delivering immersive customer experiences?

Cristina Fonseca : Immersive CX is fast becoming the new standard of CX, driven largely by what people are increasingly expecting of brands today – that they meet them where they are, under their terms, through seamless and engaging interactions. As AI becomes more cost-effective, easier to build, train and maintain, there will be wider adoption of the technology, especially generative AI with its conversational capabilities, redefining how companies engage with their customers.

Zendesk research shows that over 4 in five (84%) customers in India expect AI interactions to become more natural and human-like over time, and the ideal evolution of AI will enable customers to ask increasingly complex questions. Businesses are at a tipping point with the latest generative AI technology harnessing the power of large language models to generate such human-like conversations.

As AI is increasingly being embedded across the entire customer journey, it is also helping companies shift from reactive service to proactive while eliminating significant amounts of workload that is manual and repetitive. This can include chatbots that provide human-like conversational experiences, as well as the ability for a customer to start an interaction on one channel and seamlessly switch to another channel without the need to repeat themselves. It can also provide agents with an accurate summary of the customer’s situation and past interactions to suggest the perfect reply with the right tone that’s on brand.

The conversational capabilities and easy accessibility of generative AI gives companies the ability to respond to customers quickly, on whichever channel they are on, with the right amount of detail and context when trained well.


2. How can businesses integrate generative AI to detect customer sentiment and intent? What is the role of intent and sentiment analysis in driving personalized CX?

Cristina Fonseca : Currently, 81% of consumers in India feel most companies could be doing a better job personalizing their experiences. Unfortunately, most companies hold a narrow view of what personalization means and how to deliver it, with a vast majority of businesses in India playing catch up. What businesses need to deliver are personalized conversational experiences that work easily out-of-the-box with built-in automation when it makes sense.

And without a smart strategy for AI implementation, businesses will not only disappoint customers but also lose revenue. Large Language Models (LLMs) like the ones behind ChatGPT have the potential to provide insights into buyer behavior, including which products are most popular, which marketing channels are most effective, and what factors drive customer loyalty. These technologies can collect customer feedback, helping businesses gather crucial insights on customer sentiment, intent and pain points.

Tracking customer sentiment and intent is crucial to personalisation. We believe that 70% of customer interactions will be impacted by AI in some way, whether through deflection, automation or agent productivity. All of these require a deep understanding of what the customer wants and how the customer is feeling. Creating accurate insights and predictions require an approach to machine learning that is data-driven. When CX solutions are built upon large CX-specific datasets, the AI solution continuously learns from every customer interaction and allows companies to better assist customers with greater accuracy. This eliminates the work of manually assigning and routing inquiries freeing up team capacity and reducing operating costs.

AI can identify at-risk customers (ones in danger of churning) by conducting customer sentiment analysis to gauge intent and tone. When paired with automated routing and AI-powered workflows, the insights gained can ensure the most experienced support agents handle the tougher interactions, bypassing any self-service or chatbot workflow.

Drawing from a large set of CX-specific data, generative AI can accurately summarize customer purchase and support history for agents, who can deliver personalized and conversational experiences within minutes. Support staff can also leverage this technology for cross-selling and upselling products during support interactions.

3. How can businesses leverage ChatGPT’s conversational capabilities to improve self-service functions?

Cristina Fonseca : When customers have a problem, sometimes they want to be able to find the answers themselves. Eighty two percent of customers who have interacted with generative AI believe it will be central to discovering and exploring information. For self-service specifically, generative AI can be used to provide a conversational reply with the right information a customer needs instead of the chatbot pushing out a link to the help center or using a predefined flow and reply that needs to be manually configured. Large Language Models are very good at understanding and manipulating text, so they can be instrumental in surfacing the right information given a piece of context.

This approach can reduce the time taken to solve a customer issue and increase accuracy, compared to a traditional help center or chatbot. Businesses can get instant answers and 24/7 support using generative AI-powered bots and lean on AI to discover ways to improve self-service content to better meet customer needs.

4. Is generative AI like ChatGPT always accurate? What must businesses keep in mind while integrating ChatGPT into their CX functions?

Cristina Fonseca : We anticipate that almost all customer service will be AI-first by 2025, and in the future, AI could automate tasks to replace a majority of front line customer interactions. AI is soon expected to extend to all service needs, including proactive initiatives and preventive planning. And as businesses continue to adopt AI-driven solutions like the ones behind ChatGPT for CX functions, it’s important to not get caught up in the hype.

While adopting any AI solution, companies need to focus on applying it thoughtfully. Merely implementing a generic Large Language Model (or the equivalent to ChatGPT) for customer conversations does not always result in positive outcomes. Generative AI like ChatGPT is a great solution that provides the foundational capability, but companies that intend to use it for CX functions must ensure they are specifically fine tuned and trained for CX and designed to deliver actionable customer insights immediately. When integrated with enhanced, pre-trained bots that leverage an extensive CX database, agents can more confidently address customer queries with more personalized, industry-specific and accurate responses. Businesses can ensure transparency and accuracy by choosing solutions that provide information on the quality of AI predictions along with confidence scores, so they can know exactly what they are working with. These foundational features are key to improving customer service quality and agent efficiency, with the potential to be even more powerful when combined with LLMs.

At the end of the day, businesses looking to integrate Generative AI into their CX functions should remember that AI learns over time, and human oversight is especially crucial during the initial stages of training any AI model.


5. How can businesses balance human-automation strategies while implementing LLM-powered customer service?

Cristina Fonseca : Businesses keen on deploying generative AI-driven solutions need to bear in mind that the majority of consumers today want generative AI to improve interactions, not replace agents. In fact, 81% of consumers say having access to a human agent is critical in maintaining their trust with a business when they have trouble with AI customer service.

Having a well-thought out human-automation strategy – one that maintains human oversight throughout – will ensure AI is implemented to suit the business’ specific requirements and customer needs.

For instance, a company should set the right threshold for how complex a question generative AI can handle before handing it over to an agent. Because there will inevitably be customer queries that require a human and AI doesn’t have the necessary data to handle it.

Human touch will remain an important element for CX, so businesses should leverage AI to automate other time-consuming, manual tasks instead of replacing the entire service team with bots. The best human-automation strategies are those centered around the best utilization of resources – AI/ML should enhance or optimize this by providing assistance so that the human may employ higher value work or identify areas of improvement, thereby improving the overall experience for customers, agents and employees.

Leave a Reply