Communications Automation: Teaching Robots to Speak Our Language

Editor's note: for current information, visit the UiPath Communications Mining page.

Business runs on conversations. Think about it: how far can you get in a day's work without messaging someone?

Customers reach out when they want something. Colleagues communicate to complete tasks. Every conversation—phone, email, or in-person—means there's work to do. A problem to solve or an opportunity to gain. And every message matters when trying to deliver exceptional service.

However, after a few difficult years, business services are at a breaking point. Unprecedented volumes of digital communications keep flooding in. People feel bogged down by customer and colleague conversations at work. The average employee now spends 28% of their workweek just on email. Nearly a quarter (22%) say email volumes are the main reason they want to quit their job.

Great conversations still happen. Conversations that close deals, delight customers, and educate colleagues. But these are getting lost in a sea of follow-ups, confirmation emails, and updates. Overstretched employees miss important messages. Customers grow frustrated as waiting times increase and their requests drop to the end of the queue.

The contact challenge

Businesses need to evolve their approach to conversations. To deliver great service, businesses should focus on cutting waiting times and improving outcomes. The key is to free service teams from repetitive, monotonous requests and give them more time to focus on the most important conversations. Fewer password reset requests and more customer one-to-ones.

Automation has been crucial in freeing people from their least favorite work. Robots are everywhere—cutting waste, improving response times, and supercharging productivity. Now it's time to put them in the hands of service teams. It's estimated that half of helpdesk requests are minor in complexity, requiring one or two actions to resolve. Such requests are ideal opportunities for communications automation.

However, there's a problem. Correctly answering a question or resolving a request demands a real understanding of human language. Yet robots have long struggled to understand the nuance of our conversations.

Human conversations are freeform. We don't follow a consistent set of rules when speaking. We don't always say what we mean. Words can have completely different meanings depending on the context. A 'friendly reminder' isn't always friendly. What's more, many industries have their own terminology that doesn't make sense to outsiders.

Conversations are complicated. Chatbots are a great tool for performing simple tasks like answering FAQs. But service work is often more difficult. Complex queries need a tool that understands exactly what customers want, but also how they feel and how urgent their requests are. This tool also must learn and change as customer needs do.

Nowadays, it's easy to train a robot to do something. But communications automation requires both understanding and action. A software robot can easily perform an action off the back of a request. But you need to be certain it's the right action, every time. To date, this has required contextual and semantic intelligence that's unique to humans. Now, that’s changing.

Communications automation: transforming service

Thanks to recent advances, businesses can automate their most common requests and time-consuming conversations from beginning to end. Service and transformation leaders are achieving this by combining artificial intelligence (AI) with automation. And joining human intelligence with machine learning (ML) to improve business outcomes.

Thanks to no-code ML technologies, human service teams can now teach robots to understand the conversations happening in their business. For example, Re:infer, part of the UiPath family, makes it easy for any employee to contribute to training robots. Through a process called active learning, employees impart their knowledge and understanding of unique language and jargon used in their business.

The next stage of service automation is communications mining. Trained ML models monitor all inbound communications, in real time and across all channels—from email to online chats and calls. Then, ML models extract the most important information for analysis.

Communications mining takes a microscope to all your channels. Built-in analytics tell you exactly what's going on across your service function. Custom alerts update you on changes in service quality so you can get ahead of emerging issues.

Most importantly, communications mining creates a whole new source of information to enable automation. For the first time, processes that rely on communications can be automated, end to end.

Imagine that a customer requests an update to their address. Typically, an agent needs to read the customer’s message, and sometimes forward it to another team. Then, someone must go through that message, extract all the important information, and manually key it into the relevant systems. It's an essential process, but no one enjoys doing it and no customer will thank you for doing it well.

By combining AI and automation, however, the process becomes near instantaneous. Communications mining will scan the message and automatically extract all the key information. A software bot then performs all the updates necessary. The customer has their request fulfilled in less time, while agents are freed to focus on more valuable work.

Now, you can repeat this automation across every transactional service in the business. The impact would be huge. Picture a large, international business with hundreds of service desks and thousands of agents. Communications automation could free up millions of hours each year. Millions more hours for value creation and servicing your most important customers.

Understand and action service conversations

We'll always need humans to provide the best service. The value of communications automation is that it streamlines communications processes. Which frees up precious time for humans to focus on the conversations that really matter.

Imagine a business where no customer needs to wait days for a response. A seamless, satisfying service experience for employees and customers. Communications automation is another step on the road to the fully automated enterprise™.

I had the pleasure of introducing Re:infer—the newest member of the UiPath family—at FORWARD 5 this year. If you’d like to see Re:infer and communications mining in action, you can watch the replay of “Meet Re:infer a UiPath Company,” my session at FORWARD 5. To access the recording, select “Watch the recordings” on the FORWARD 5 site. Access to the recording is free.