The state of play for Artificial Intelligence (AI) in the contact center 

There is no doubt that when Open AI’s, ChatGPT, burst into the public consciousness in 2022 things changed forever. 

Instead of hiding away in some spurious black box, espoused by technology vendors in generic language, an example of AI was finally understandable, relatable, and crucially, accessible. 

A human could interact with ChatGPT in a human way and in return, receive a human-like answer. 

Game changed. AI was no longer purely the pursuit of the technology zealots. 

And yet, AI has obviously been around and impacting the world much longer than this (in fact it can trace its roots back to the 1950s). The difference being the pace and breadth of change. The volume of AI models, for example, would not have been countenanced by many ten, or even five, years ago. 

But what does this mean for contact center operations? 

AI will open up a wide world of opportunity for the contact center and should evolve how we manage customer interactions at a faster pace than ever before. 

That sounds like a bold statement, but I’ve already seen a lot of beneficial use cases, and it’s still early days. 

In this article, I’ll run through some of the key considerations you must make if undertaking an AI-driven project in your contact center or getting into bed with a technology vendor lauding the value of AI. 

Why AI should be used in the contact center 

Firstly, let’s examine the opportunity. 

The public use of AI has exploded in the last 12 months with people creating recipes, poems, web copy, you name it. The interaction point has largely been via Large Language Models, such as ChatGPT. 

This is largely reflective in the contact center as well, given the LLMs ease of adoption and use. We can use ChatGPT to summarize calls, for example. 

There are other AI use cases at work though, and I’ll examine a few below. 

1. Automate interactions – We’ve all interacted with a chatbot before and I’m sure, like me, you’ve had mixed experiences. For automating routine tasks, chatbots can be extremely useful. Simple interactions where humans have no interest in talking to another human is the most effective use case. As things get more complex, then typically you need to hand off to a human or let a human lead from the outset. Balance is key. 

2. Predictive analytics – AI can help operational teams and agents make better decisions through predicting customer behaviors and analyzing sentiment. Here at Awaken we specialize in intent analysis, understanding what the customer wants live in the moment and helping the agent find the necessary information to service that query. 

3. Save time – An obvious one. AI can help automate repetitive tasks, such as looking up information in other systems, analysis of customer information and profiles, summarizing calls etc. By making interactions more efficient you can reallocate resources or save money. 

4. Improve quality – This is a really important one. AI has the ability to help personalize customer interactions. It can also help analyze vast quantities of interaction data to understand where processes are falling over. Which can then be fixed. Crucially, it can also improve the audio quality and audio output (be that a recording / transcription etc.) Which all pulls together to equal better quality interactions and happier customers. 

The exciting part is we’re just getting started. There has been a leap in accessibility of AI and the beneficial use cases it supports will only continue to grow as technology vendors get to grips with it more. 

Security should be a credible concern 

In modern society, when we take a technological leap forward, the buzz and hype often outrun the pragmatics of the advancement. But it does quickly catch up and re-align. 

We’ve seen this with LLMs, such as ChatGPT. Everyone was quick to adopt it, we saw publicly how some organizations sent it data it shouldn’t have, and then things almost locked down. 

Organizations are being far more rigorous with their approach to LLMs as a result. But that’s not to say the opportunity has disappeared. 

What we’re seeing now is almost the third phase of AI adoption (in relation to LLMs) in which organizations have satisfied their infosec insecurities and are once again approaching AI to solve their issues. 

We’re also seeing a broad desire for use cases such as live data redaction, an obvious one given the increase in regulatory scrutiny across all markets, and something AI makes far easier than ever before. 

Like balancing the opportunity, the secret lies in your ability to adopt the appropriate technology based on the use case, organizations must also satisfy their security responsibilities.  

That means mapping processes, understanding what data is fed where and how it will be used by the AI models. 

ChatGPT caught the eye purely because it was so publicly adopted. I wonder if the AI models hidden deeper in some of the industry’s major tech infrastructure have received the same level of attention? 

How is Awaken using artificial intelligence? 

Awaken has three core products that blend together to provide one seamless contact center solution. 

We have agent guidance, conversational analytics, and now, real-time agent assistance. 

Artificial Intelligence supports all of them. 

Whether it’s analyzing customer and agent sentiment in real time, summarizing interactions for easy consumption, or redacting information, such as PII, without AI our portfolio would certainly not be as beneficial to customers. 

If you’d like to discuss your contact center operations, and how AI can support your goals, then get in touch today.