At ROGER365.io, we always like to keep our finger on the pulse of innovations in enterprise communications, often by attending major industry events and talking with experts across the sector.
We had such an opportunity at Enterprise Connect 2024 - the #1 event for the communications and contact sector.
In this blog, we share the biggest developments in technology and groundbreaking new trends from Enterprise Connect 2024.
The greatest trend in enterprise communications
The biggest news (by far) from Enterprise Connect 2024 was the rapid proliferation of AI in contact centers and enterprise communications. This trend is dominating the sector.
Whenever new technology enters the scene, people immediately start to imagine what it could achieve. High expectations often give way to a ‘trough of disillusionment’ (a term coined by Gartner) when the technological reality struggles to keep pace and deliver on these expectations.
Eventually, however, progress catches up, and increased technological maturity allows innovations to fall into sync with our understanding of their capabilities.
What’s the real potential of AI?
For businesses looking to gain maximum productivity from new technology, the best strategy is to stay close to the ‘racing line’ – that sweet spot where you get the most value from the least risk.
Or, in other words, using proven use cases as the basis of tech investment decisions, while leaving the door open to new capabilities as they become available (and reliable).
So, what are the current uses of AI in Contact Centers, and what capabilities are likely to follow in the near future?
How AI is already improving contact center experiences
Better customer experiences from AI-powered chatbots
Sometimes, getting the right answer out of a chatbot can feel like trying to get blood from a stone – and this can be frustrating. Thankfully, AI has particular value in upgrading simple ‘FAQ-style’ chatbots into intelligent conversational chatbots that can detect customer intent.
By training the AI on your own documentation (such as transcripts, brand books, business rules and other customer service records), an AI chatbot can provide better assistance that truly matches the customer query.
Using the power of Large Language Models (LLMs) and Natural Language Processing (NLP), a conversational AI chatbot can give customers context-driven solutions in seconds and reduce the need to redirect to human agents.
As a result, customers get rapid answers to their most common queries, and the overall experience is better - because they don’t need to spend time trying to coax a satisfactory answer from a ‘dumb’ chatbot that doesn’t really understand the query.
AI voicebots in contact centers
Another proven use case for AI in contact centers is with voicebots. These use a speech-to-text engine to bridge the gap between voice calls and conversational AI, while leveraging the AI’s capability to understand intent.
Voicebots allow your customers to rapidly access AI-powered solutions to their queries via a regular voice call. This is a huge step-up from keyword-based voicebots, and it can also enable you to offer multilingual support across all channels.
Deeper analytics and insights from data with machine learning
Advances in machine learning are already positively impacting the potential value of customer experience data. As these become more sophisticated and customizable via conversational AI interfaces, it will become easier to pick out hidden patterns in customer behavior, discover common themes and topics, and understand what is needed to improve.
In the near future, data analytics will become further enhanced as AI produces entirely new data streams, including advanced sentiment analysis and cross-channel that connect the dots across more touchpoints.
AI-powered Agent Assistance in Teams Contact Centers
Agent assistance (or ‘Agent Assist’) is one area where AI can massively increase productivity. It also has the least risk potential because it isn’t directly customer-facing.
Teams Contact Centers already have a lot to gain thanks to the continued development of Microsoft Copilot as a powerful AI-powered assistant for customer service agents. (Read more about this below).
The AI-powered contact center of the near future
Real-time voice call transcription in Teams
Very soon, Teams will support transcription for regular voice calls in real-time. This means that each interaction can be documented in granular detail and accompanied by an automated AI-generated summary of the conversation.
As a result, more knowledge of the customer is captured, and agents are liberated from the drudgery of typing these up and saving them in the CRM manually.
Instead, it all happens automatically.
Enhanced sentiment analysis (real-time)
Sentiment analysis is not a new thing, but it can be imprecise and often arrives too late to be useful. What you need is a clear indication in real-time when a customer query isn’t going the way it should. With AI-powered sentiment analysis already on the horizon, you can get a clear overview of customer experiences as they happen, with the AI picking up on more subtle signals like crosstalk, negation, tone shift, and keywords. This means that Supervisors can whisper advice to embattled agents, or barge-in and take the lead.
More secure authentication with voiceprints
Customers must often authenticate their identity with ‘secret’ information – but what if this data is compromised? In this scenario, any caller could state the required information over a voice call and gain the same access as the real customer. However, thanks to AI’s ability to recognize voiceprints, the caller needs to have both the information, and the customer’s voice. With an AI-analyzed voiceprint, authentication becomes far more secure (and convenient).
Fraud detection
But what if the fraudster uses an AI voice generator to fake the voiceprint?
This can certainly be done using currently available technology and publicly available voice samples from social media posts. Indeed, this tactic has been used to defraud individuals by impersonating people they know. However, once again AI comes to the rescue because there are already tools available for detecting fake AI voiceprints using telltale signatures human ears cannot detect. Likewise, other unusual behaviors can be flagged, and fraud more easily detected.
This means that you can offer a higher level of authentication to customers.
First-class AI agent assistance
As we touched upon above, Microsoft Teams Contact Centers can already make use of Microsoft Copilot as an agent assistance/productivity tool. However, this capability is set to greatly improve in the coming years.
As conversational AI becomes more sophisticated, it can be trained on your data and documentation to offer suitable suggestions to agents during a call.
By listening-in on the conversation, the AI assistant can bring up relevant courses of action, links, information, and documentation that can help the agent solve queries without needing to search for them.
Furthermore, an AI assistant can draft responses or templates, and create summaries that are automatically saved to the CRM afterwards. With the right setup, an AI assistant can also connect the current conversation with all previous interactions, so the full context is brought into view.
Accelerated agent training
By training conversational AI on documents like customer interaction transcripts, business rules, and employee guidelines, your agents can be trained faster and better. This method means there’s a reduced need for experienced agents to shadow new colleagues.
The AI can act as a ‘real customer,’ simulating common queries with ‘real’ conversations across different channels. This can dynamically adjust to the most common issues of the day, meaning that your training program always reflects the required skill set.
Furthermore, new agents can be given enhanced assistance that suggests the best verbal responses as well as recommended actions, processes, and documents.
With an AI-powered training program, your new agents can confidently get started (and faster), with the benefit of all your collected expertise and experience.
How to avoid risks from using AI in contact centers
The #1 risk of AI is that customers aren’t satisfied with the AI, and want to speak to a real person. This means that you must make it as easy as possible to redirect queries at the slightest signal that your AI isn’t sufficient for their needs.
Next, there’s a whole collection of risks that come from bad implementation.
When AI is set up correctly, there should be minimal risk of problems – so it’s vital that best practices are adhered to. For instance, to deliver value, AI must have access to the right data and resources. This will require careful oversight to ensure that integrations are complete, securely configured, and well-maintained. You also need to give AI access to only the right resources, and create rules that ensure that sensitive or private information isn’t compromised.
Training is also vital, for both the AI and your staff. For people to get the most value from AI tools, they need adequate training in how to use them effectively and securely.
Security should always be at the forefront of your mind. Sadly, there are always some people who will try to exploit or manipulate an AI chatbot or voicebot into saying something or releasing information it shouldn’t. For this reason, you should pay close attention to which data you give access to and how you can prevent malicious actions.
Access control is one part of the solution; another is the implementation of guardrails. Think about ways you can detect ‘risky’ topics, unusual behaviors, or weird question series, and create rules that redirect suspect interactions to a real person.
Getting started with AI in your contact center
Microsoft Copilot is one of the best AI-powered tools already available, but new solutions and vendors are entering the market each day. Every one of them seeks to create a tantalizing, unique offer - and makes big promises about what it can achieve.
This might leave you wondering where to start. The answer is to look at use cases where AI has already established a proven value, and build it up from there.
Our view is that agent assistance is the best place to start, as it offers the greatest gains with minimal risks. Customers will immediately experience the benefits of better, more efficient service, and agents are instantly unburdened from all those tiresome and repetitive tasks.
With a ROGER365.io Teams Contact Center it’s easy to take an incremental approach, by integrating each capability in a modular way. This allows your organization to get to grips with each new improvement, and to learn how to leverage it to deliver the best experiences to customers.
Want to learn more about implementing AI with your contact center? Get in touch, and let’s discuss your specific situation and needs.
FAQs
How is AI used in contact centers?
AI is already being used in contact centers to improve customer experiences by analyzing data from interactions for insights, and by resolving many queries without requiring a human agent. It allows your contact center to provide more convenient self-help options, which increases your overall availability.
How to use AI in customer service?
You should take care to use AI in customer service in a way that doesn’t compromise the quality of your service: give the option to redirect a query to an agent, limit access to sensitive data, and make sure the AI is trained with your own transcripts and documents. When implemented properly, AI can be used to provide customer service via chatbots, voicebots, and through multiple channels including email, WhatsApp, and voice.
What is AI agent assist?
AI Agent Assist is an agent assistance capability that is powered by AI. There are several vendors already in this field, however in coming years AI agent assistance will become more sophisticated and capable, as well as highly customizable to your situation.
What’s the future of AI in contact centers?
AI has a bright future in contact centers, especially for providing convenient answers in a fully automated way via text, voice, and through practically every channel. It will continue to take the burden off human agents by taking care of the most repetitive tasks, and make their job easier and more fulfilling.
Register for webinar
Download whitepaper
