How machine learning is changing business communications

The business communications landscape has changed dramatically in the last twenty years, thanks to Voice-Over-Internet-Protocol (VoIP) and the rise of the internet. Telephone networks are no longer tied to a wired landline network. Instead, your VoIP phone system can be operated from various IP addresses, providing flexibility for workers and devices. Machine learning is further transforming that landscape.

Gartner, in their report, Top 10 Strategic Technology Trends for 2020, state that AI and machine learning are increasingly used to make decisions in place of humans. In this article, we dig into this trend to focus on how machine learning is changing business communications.

SEE ALSO: 4 tools that solve 4 problems of remote work

What is machine learning?

Machine learning is a subset of Artificial Intelligence (AI). Artificial intelligence is based on computer algorithms that can process statistics quickly. With AI, it is possible to perform a task or series of tasks and run them repeatedly. It is ‘intelligent’ because it performs quite complex duties.

Machine learning takes this one step further by providing the program with a model. The program processes the data to reach conclusions. Machine learning can be broken down into three groups:

• Supervised machine learning: takes already known sets of data and then responds to queries or tasks.
• Unsupervised machine learning: the program finds new patterns and forms a model of its own.
• Reinforcement learning: through a process of trial and error, the algorithm will make specific decisions.

Machine learning already has a range of real-world applications. It powers the way that Google analyses and powers search results, provides the external input for self-driving cars, and more.

To recap, machine learning is not artificial intelligence, though it is a subset. It is not training computers to roll out responses to previously programmed queries. Machine learning is about creating programs that, given a series of data, can make intelligent, smart calculations, and that can learn to iterate and improve their responses.

As you might expect, machine learning is being applied extensively in real-world situations. The telecommunications sector is no exception. It’s being used for VoIP, virtual assistants, and more. In the next section of this guide, we’ll focus on a few notable developments.

Microsoft’s vision for intelligent communications

Microsoft led the way in unified communications. “The company was UC before UC was UC” quipped Robert Ballecer at Enterprise Connect in Florida earlier this year when interviewing Scott van Vliet, Corporate Vice President for Microsoft Teams. Microsoft Teams has brought “together all the aspects of communications and collaboration, whether it is chat, calling, meetings, and the office suite of apps,” says Scott.

Over 500,000 organizations using Microsoft Teams. Microsoft themselves have 180,000 team members using the cloud-based software. Scott is a keen proponent of the innovations the team is making in video quality. Looking forward, they plan ‘intelligent’ video background replacements. Both features use machine learning.

Microsoft announced a vision for ‘intelligent communications’ on their blog back in September 2017. The idea was “to transform calling and meeting experiences for people and organizations around the world. To achieve our vision, we are bringing comprehensive calling and meeting capabilities into Teams, and infusing those experiences with intelligence.”

One of the significant features discussed in the keynote speech by Bob Davis, Corporate Vice President of Office 365 Engineering at Microsoft, at Enterprise Connect was the addition of Microsoft Graph to Teams. Microsoft describes Graph as “the gateway to data and intelligence in Microsoft 365”. GetVoIP cites Davis as calling it “a powerful brain that connects to data signals coming from every email, every file, every message sent within Office 365.”

Source: Microsoft

Examples of what Graph can do within a team include finding someone by name, and not just by finding a simple search match for a phrase. Graph will calculate which person with that name you most commonly deal with or which person with that name is in your team or channel.

Graph goes one step further. If you don’t know the name of the person you are looking for, but you need someone with expertise in, say, ‘chatbots’ Graph can calculate who in the organization you are most likely to need. The program bases its calculation on who posts, chats, or emails most frequently about ‘chatbots.’

These features scratch the surface of the possibilities. Graph can provide virtual assistance for meetings, recording, transcribing, and even filing meeting notes. You can also search for a particular keyword, or speaker and Graph will take you to that timestamp in the recording of the meeting.

There appears to be a common thread in the machine learning features list in the improvement of meetings. Diane Chaleff, Office of the CTO for Google Cloud Suite, also spoke at Enterprise Connect on Google’s plans for machine learning and their vision for future communications.

Google’s plans for the future of communications

Diane highlighted the time-saving features of an intelligent virtual assistant. A bot that could scan the calendars of meeting attendees and set up a suitable appointment after voice activation, e.g., ‘Set up a meeting with John,’ a simple act such as this could save employees up to an hour a week. More complex tasks can, of course, be completed.

Voice calls are already customizable in Google’s G Suite, allowing a business to assign & port calls or review billing in a simple console. Google’s Voice already uses AI to filter out spam calls, integrates with Hangouts and Calendar, and transcribes any voicemail for you. Another interesting product being developed by Google is Jamboard.

Google Jamboard is an interactive business whiteboard. Google has plans to improve the machine learning behind the digital whiteboard to do things like taking notes. These notes will accurately transcribe conversations, even when there are multiple users. Jamboard will be able to take notes, save them, and potentially also send them to the meeting attendees.

Chaleff adds that “it’s critical to remember what doesn’t change. The key to all forms of communication is to maximize human ingenuity. Get the junk stuff out of the way and leverage machines so people can do what they do best — be creative.”

There is a mindset shift in this direction. It moves us away from scary predictions about AI and refocuses us on what artificial intelligence can do to help us.

SEE ALSO: 2020 State of Enterprise Machine Learning: Cost reduction is the top priority

What AI can do to help business communications

The improvements in machine learning do sound like sci-fi, but both data and people are important to business communications. Technology has the potential to improve the bond between people and information, allowing for improved qualification of sales leads, better profiling of customers, and enhanced customer service.

Modern business communications produce vast volumes of data. Cutting edge developments like machine learning allows us to process larger quantities of data faster, for example, by recording and storing all conference calls. Recording, storing, and processing call data allows a business to identify patterns for increased productivity, identifying trends in team and customer communications that need to change or where they are working.

Machine learning can also help businesses cope with the ‘always on’ expectation of customers. Customers expect an immediate answer to a phone call, and employees are expected to be by the laptop for conference calls wherever they are working. By handling emails with intelligent automated responses, an employee is relieved of at least part of their email influx.

Machine learning might also soon be able to tell us whether the customer prefers a call or meeting, and when to speak vs. when to listen to a customer on the call. The data will help customer service teams work out what time of day teams need to be on calls and how many employees we need at any one time. Anything that requires the processing of volumes of data to identify patterns could, in theory, be assisted by machine learning.

Conclusion

So, there we have it, two major players in the tech field are working towards improving business communications through assistants and meeting software that is informed by machine learning. Each of the technology giants already has VoIP call functionality built into packages; Voice for Google’s G Suite and Phone System in Microsoft 365.

They are both working towards improving systems using machine learning, and the results will change the way businesses communicate in significant ways.

There is much to look forward to in terms of discovering how machine learning can help business communications to improve. Twenty years ago, we envisaged machine learning as a scary technology that was a futuristic sci-fi dream. Today it is fast becoming a realistic and practical tool for business communication.

The post How machine learning is changing business communications appeared first on JAXenter.