WebRTC Live 88: How WebRTC and LLMs Combine to Lift Human Conversation Performance

As language learning models, such as Symbl.ai’s Nebula LLM, continue to evolve, there is greater potential to revolutionize real-time interactions. 

On the February 21, 2024 episode of WebRTC Live, Arin welcomed Dan Nordale of Symbl.ai.  Symbl uses advanced LLM technology to help businesses analyze large-scale multi-party conversations and ultimately make better-informed decisions. In the episode, they dive into four ways this type of Generative AI is elevating live communication: AI Agents and AI Agent Assist in the contact center, intelligent meetings, and enhanced sales training.

Dan is a revenue leader with a passion for bringing Conversational AI technologies to communication platform developers. With a focus on customer-centric strategies, he has led the go-to-market organizations of rapidly growing software companies.

This episode also featured the debut of a monthly WebRTC Trends segment with Tsahi Levent-Levi.

Scroll down for Key Insights and Episode Highlights, or head straight to: 

Watch Episode 88!

Key Insights from WebRTC Live Episode 88

LLMs can help improve communication and decision-making in customer interactions. One of the most beneficial use cases of LLMs is call centers, where they can help customer service reps make better decisions and provide better service in real time. 

Dan says, “Can we use an LLM to deeply understand the context midpoint in a conversation based on what’s said, and can we return very bespoke, specific, context-aware guidance that understands what’s been said? It understands the context around that type of interaction. It understands the sentiment of both the customer prospect and the employee, and it probably understands the psychological underpinnings of what’s going on around that type of interaction, and it responds accordingly, and it responds in a timeframe that allows the seller, allows the customer service agent to use that  intelligence in to deliver better outcomes.”

Combining automation with human oversight will improve customer satisfaction. Automation streamlines tasks, but humans handle complex matters and add a personal touch to interactions. That’s why they work best when combined.

Dan says, “I’m pretty receptive to automation in many cases because it’s efficient. And so I think that there’s huge potential here, but we see it going further. We’re seeing this around prospecting calls.”

Arin adds, “Having a successful prospecting interaction and then transferring to a human agent to close the deal to talk to finer details or in a use case, like we showed in this video of kind of booking some travel, just having a human in the loop agent who then gets on with you and confirms before you actually take the credit card or make the booking that the dates and the locations and everything are right and all that sort of stuff. So, there’s a lot of different ways that we could still combine the agents and the human agents and going sort of seamlessly in and out of the conversations and keeping that context across them.”

AI is a powerful tool that can improve our capabilities and make us more efficient. Instead of fearing AI, we should learn to embrace it. While it’s true that some jobs might change moving forward, there’s no reason to worry. We can work together with advanced technology to be more efficient at our jobs and improve our capabilities.

Dan sums it up nicely by saying, “I think it’s one of the most exciting ways where this technology, which we often perceive as replacing people, can be employed to make people tremendously more effective in what they’re doing. 

Episode 88 Highlights

We’re expanding the limits of generative AI

The future is certainly bright for generative AI and customer interactions. AI can help us be more efficient while at the same delivering richer and more valuable experiences for the customers. Dan explains, “We have a lot of really interesting customers who are really stretching the limits of how you use AI and how does that assist both your employees to make them better and how does it assist your conversations, these durable, long-term, long-ranging conversations that you’re having with consumers and how do we make those rich to serve the consumers intends and needs and also make those fruitful for organizations as well. “

AI can help improve productivity and provide greater value to live interactions

By focusing on the end goals of meetings, such as identifying and distributing work items effectively, AI technologies can greatly enhance productivity and outcomes, offering unique value in meeting contexts. Dan says, “If we start with the end in mind, we can deliver some extremely useful productivity-enhancing and really outcome-enhancing capabilities in the context of the meeting. And that’s where we think the excitement, and really unique value, is going to be delivered in AI relative to meetings.”

Technology is revolutionizing real-time support in customer interactions

As technology evolves, it becomes more helpful for call center agents and salespeople to respond to customer needs more swiftly and effectively. Dan says, “Largely, it’s understanding what’s being said in the conversation and then using, in some cases, it’s just static assistance, but more and more, what we’re talking about is generative assistance to an agent based on what’s been said, and it returns to an agent in real-time, sub-second, and it allows them to understand what is the right resolution path for this customer what is the right way for me to get to my goal in a sales conversation based on where I am.”

Automation allows for continuous improvement

AI isn’t here to replace us but make us better at what we do. We can use automation to consistently improve our skills and learn from every single live interaction right there on the spot. Dan explains, “We can automate the grading of performance in every single interaction. So this can be live. This could be on every call. We can say, man, great. You did great in these three elements and engaged the customer well. You followed the process well. You set goals. You established next steps. You did a lot of things really well. Next time, you could improve in these areas … ”

Up Next! WebRTC Live Episode 89
with Tsahi Levent-Levi

Wednesday, March 13 at 12:30 pm Eastern


Recent Blog Posts