Skip to main content

Zendesk Head of AI on the current value add of generative AI and future applications for CX

Following her debut on the main stage at Zendesk Relate, we caught up with Cristina Fonseca on all things AI

Published June 21, 2023

A lot has happened since we last spoke. At Relate, the company announced Zendesk AI. What did you hear from customers in response to this big release?
Customers are ready and willing to accelerate their adoption of AI and machine learning solutions. In talking with customers, I found that many are excited about the latest developments in the AI world, but have been patiently (or not so patiently) waiting to use AI tools from trusted vendors they already have a strong relationship with. Zendesk AI builds on our existing AI capabilities and continues to deliver value from day one.

AI is top of mind for all leaders. Can you explain the competitive edge for Zendesk AI compared to solutions in the market?
We have been building an AI solution that truly speaks service. Our core models are proprietary and have been intentionally built for customer experience (CX) use cases, based on what we believe to be the world’s most comprehensive CX-specific dataset.

Here’s an example of how Zendesk AI provides that precise level of detail: when the ticket sentiment reveals a customer is frustrated with a bot, that will come out as ‘negative’ because of how the sentiment model is calibrated. This level of accuracy and explainability can only be achieved with models that have been trained in house. Our intent detection mechanism is as granular as needed for brands to be able to act on each type of request with a personalized touch.

Here’s another example for the retail industry in which a customer is asking about order status. We can not only detect that the customer has a question about order status, but we can go one level deeper, distinguishing between asking about the order tracking number, if the order has been shipped, when the order is going to arrive, or help finding the order tracking number. Our customers are excited for these intent and sentiment analysis capabilities – looking ahead, they can take the conversation to the next level by automating interactions.

The applications of AI are very real for retailers, agents and shoppers. How else are brands in the e-commerce space applying AI in CX?
One of the reasons why there’s so much potential in AI for customer service is due to the manual and repetitive tasks that are currently performed by humans, but can be done by AI. For example, New Zealand’s most popular and longest running online shopping sites was dealing with a high volume of requests; and they had a dedicated person spending their entire day manually triaging tickets. They were able to turn on Zendesk AI and within hours, fully automate common issues, and route the right tickets to the right agent. This freed up one full time agent to work on more fulfilling tasks. And their customers are getting resolutions faster while maintaining their high quality standards. Other retail use cases include leveraging AI-powered customer intent data together with analytics showing pages viewed or past interactions to recommend personalized product offerings and improve conversion and retention, which leads to increased customer loyalty.

Can you share some everyday examples of how other industries benefit from using AI?
Let’s take a look at the financial services industry where timeliness is not only important to customers, but required to avoid sanctions and fines in certain regions. Companies can use AI-powered intent and sentiment detection to categorize compliance-related tickets, then use time-based automations to prioritize and assign tickets to the right agent group. Reminders can also be sent to agents based on how long the request has been waiting in the queue to ensure that tickets are addressed within the legally required time limit. This will help ensure compliance and reduce regulatory fines due to customers escalating complaints to regulatory bodies – for example, credit card disputes need to be investigated and resolved within the prescribed timeline. Additionally, GDPR or CCPA requests such as right to access or right to erasure need to be responded to within an allotted time otherwise companies can be subject to sanctions and fines.

What’s the next big opportunity for AI?
These recent advancements equal massive opportunity, like ways to deliver even better CX. AI is becoming so deeply integrated with databases and systems that it can completely make decisions and fully automate actions – issue refunds, build coupon codes, understand what it takes to deliver intelligent CX.

From your perspective, how will AI change CX in the coming years?
Customer inquiries may stay the same – making returns, changing flights, etc. – but we will be able to resolve them faster as AI evolves at a rapid pace. At Zendesk, we remain loyal to solving the problems that our customers’ service teams are facing, such as manual and repetitive work, in order for these CX interactions to be easy, frictionless, and helpful.

Our teams are working to extend our AI capabilities to do entity detection and extraction to automate even more. In addition, we are working to have AI deeply integrated into CX workflows so companies can use AI to personalize interactions in every channel for a consistent experience. Lastly, we are leveraging generative AI to provide personalized answers to customers faster.

For more information on Zendesk AI, visit our website here.