Talkdesk launched two new generative AI (Gen AI) tools meant to address the challenges associated with traditional interactive voice responses.
We’ve all faced the frustration of interactive voice response (IVR), which often fails to address specific problems and can be highly irritating. IVR systems make getting through to an agent difficult, leading to a poor customer experience (CX). At Customer Contact Week (CCW) Las Vegas 2024, Talkdesk launched two new generative AI (Gen AI) tools that address the challenges associated with traditional IVRs.
Talkdesk’s Gen AI-powered Navigator
The first tool, Navigator, leverages Gen AI to provide intuitive and context-aware customer interactions across digital and voice channels. Customers will encounter a continuous conversational IVR experience with Navigator. It manages and prioritizes inquiries, understands customer requests in natural language, and offers personalized responses without using complex IVR menus. For example, Navigator can access a customer’s account data, recognize their open claim, identify a claims representative, and direct them to the right person.
Talkdesk Navigator is similar to other offerings from a conversational and hyper-personalization perspective but takes it further for CX and CCaaS use cases. And, Talkdesk’s platform includes voice, digital and contact center capabilities.
Navigator offers significant time savings for contact center administrators by simplifying handling different customer intents. Administrators only need to add a single intent topic like “schedule appointment,” and Navigator’s large language model (LLM) automatically interprets all possible phrases once the intent is set. This lowers abandonment rates and reduces the average handle times typically associated with IVRs. It also reduces the workload for administrators who would otherwise spend time updating intent phrases.
“Navigator provides seamless AI-powered customer journeys. It facilitates natural conversations and lets customers express themselves in their voice without being forced into a frustrating menu. It also empowers administrators with AI conversations and eliminates all that tedious setup,” said Kevin McNulty, senior director of product marketing at Talkdesk.
Furthermore, Navigator supports multiple languages. If a person asks to schedule an appointment in French, Navigator will recognize and process the request in the original language.
All these capabilities are integrated within Talkdesk Studio, where Navigator is now an additional field. Navigator uses Studio’s pre-existing workflows and industry-specific configurations. According to McNulty, early previews with customers have shown impressive results, indicating significant improvements in call abandonment rates, customer satisfaction and ease of use although he did not provide specifics.
“It’s the easiest setup – it takes about 30 seconds to a minute. The only setup required is creating the workflows. But beyond that, it’s just turning on the functionality,” said McNulty.
Talkdesk’s Gen AI-powered Mood Insights
The second tool unveiled at CCW is Mood Insights, which is available within Talkdesk Interaction Analytics. It enhances sentiment analysis by using Gen AI to identify customer pain points and uncover actionable insights from post-call data. Typically, sentiment analysis categorizes interactions into positive, negative, or neutral at the end of a call. Mood Insights, however, evaluate emotions like annoyance, gratification, and joy throughout the interaction. This provides a comprehensive view of changes in customer sentiment.
Leveraging generative AI, Talkdesk can get a fine-grained understanding of what is happening throughout an interaction. If a customer is calling in about a billing error, instead of asking them at the end about their sentiment (positive, negative, neutral) – or even capturing that automatically without a survey question – Mood Insights enables the business to see that they started the call frustrated, but then a specific action taken by the agent shifted the mood to thankful.
If the call ended optimistically, there are much more precise ways of determining whether what the agent did was effective. If so, it can be implemented programmatically to affect future customer interactions, whether live or autonomously.
This is superior to survey-based sentiments, which are usually biased because people only complete them if they are super happy or unhappy. Throughout the interaction, this automated, fine-grained mood analysis gives brands more precise insights to improve CX. Also, implementation isn’t a complex mapping exercise; this happens with no configuration at all, and the AI model can be trained using Talkdesk AI Trainer.
“With these two launches, we feel this is an innovative use of gen AI. They’re starting to lead people down a path towards hyper-personalized experiences, which take into account things like mood, emotions, and are more empathetic across the entire interaction,” said McNulty.
The tools leverage gen AI to capture a wide range of feelings beyond traditional sentiment categories. A deeper understanding of how customers are feeling leads to more compassionate and responsive agents, and ultimately, fosters greater customer loyalty and satisfaction. This should be a natural progression for any company looking to transform the customer journey.
