NICE has given its robot a name — ‘the NICE employee virtual attendant’ now goes by the name of NEVA.
And NEVA and NICE technology helps in augmenting people, or so Gareth Hole, from NICE told us.
It boils down to the difference between attended and unattended robots. “Maybe”, Gareth estimates “with an organisation, 10 to 20% of the activities can be automated, end to end. But the world’s a complex place.” And when Gareth talks about augmenting people, he is thinking about the other 80 to 90%.
RPA: we take a look at UiPath, Blue Prism and Automation Anywhere
Robotics process automation, or RPA, is becoming big. And this month, RPA is our main theme. Information Age talks to three of the top players: Blue Prism, UiPath and Automation Anywhere. So here we begin RPA month by comparing and contrasting. Read here
Take a call centre. This call centre gets inundated with calls on a certain theme at certain times of the year. During these times, it may need to pull staff in from other areas, but who lack specialist knowledge.
NICE supported a public sector organisation with precisely that challenge.
“They found that they had to double the size of the contact centre during those times.” But this created challenges,
“Often, staff would have to look at a couple of dozen different screens to work out what information they needed to progress. Often, there were multiple issues that they needed to look at.” Human nature being what it is, this meant that they “might miss things inadvertently.”
“We put a virtual assistant on the desktop.” It was a quick turnaround, “it went live after about six weeks to 4,000 people.”
The new system meant that the software gathered all the information from all the different screens very quickly, interpreted it and gave feedback to the person taking the customer call, saying ‘this is fine, this is fine, here are a couple of things you want to talk to the customer about.‘“
Prior to the new software being applied, a seven-minute call to a contact centre may have entailed three minutes gathering key information and not always accurately. “So, no sooner did a customer finally get through to an advisor, they would be put on hold. Whereas with the virtual assistant, it was gathering all that information — I think the median time was about 16 seconds, but the first bits of information came back within two or three seconds. In short, the advisor could talk to the customer as the robot gathered the information.”
So they could literally just carry on the conversation.”
The future of robotics: A convergence of the physical and digital
Who is NICE?
NICE has been around for around 30 years. In the early days they focused on recording calls in contact centres. From there, the company “looked at ways of extracting value from all the data that was collected, and the data grew with new channels, chat, email and SMS.”
The route into AI, RPA and automation grew from that.
Today NICE is listed on the NASDAQ, they have a workforce of around 6,000 globally with about 30 different product lines.
Augmenting people, rather than automating end to end
And that origin seems to give NICE its unique selling point. Other companies, operating in this space often started automating back-office processes — finance and HR, for example — looking at processes that could be automated end to end effectively. For NICE it is about supporting workers in their day to day jobs.
“We started with real-time process optimisation in contact centres’ front office, helping people with work, not necessarily looking for processes we could automate end to end. If you’re able to augment people as they’re working and get them to focus on the right areas where they can add value, such as having a conversation or understanding the nuances of customer history, then our view is that you can have robots and people working together.
“That is more powerful than just trying to automate things.”
All in the mix: AI is about augmentation, not just automation
Not just about saving time
The benefit is not just about saving time. Returning to the example of the call centre, Gareth said “It was a much easier experience for the customer and for the people taking the calls.” And because each call could be completed more quickly, the contact centre could field more calls, meaning customers didn’t have to wait so long for the initial reply.”
He gave as another example, virtual assistants helping insurance claims. The virtual assistant can gather all the documentation and can put it in the right kind of format, providing the contact centre employee with a guide, thereby focusing time and effort. “We’ve had examples where you can knock an hour off a 2-hour case and get people doing the more interesting work.”
In other areas, we are also looking at adding artificial intelligence on top of that to help with some of the decisions. So we are looking in healthcare at triage, for example.”
Closing gap with customer
“Imagine you’re going to go to work in a contact centre tomorrow, for example, to do your 8-hour day and take 70 phone call/meetings with customers, with each meeting expected to take seven or eight minutes. That’s a lot of meetings, and in most cases the customer will be much better informed than they are. That’s not a great situation to be in.
“So our brief was to level the playing field, with the virtual assistant providing support, such that when a customer did phone, the employees knew as much about that customer as possible, what they’d done before, what they might be calling about, and then, within that conversation, guide it. So they weren’t at a disadvantage.”
RPA and its expansion into AI: Driving a new era of business and IT alignment
AI and RPA in harmony
Data! These days data is, well, it is supposedly the next oil. And NICE reckons that AI and RPA can provide a new way to generate data.
“When someone is interacting with systems on a desktop, tapping away for eight hours a day, there’s a lot of interaction, a lot of mouse clicks, keyboard entries, etcetera. We can turn each one of those events into an item of data.
“We reckon, with someone working eight hours a day, that’s probably 8,000 to 10,000 bits of data concerning how they interact with systems on the desktop. Once you get about 20 people doing the same thing for a couple of weeks, you get to the level of data where actually you can start to use machine learning, and text analytics, to spot patterns in what they’re doing with these systems. This means you can help with what people are doing on their desktop — augmenting people.
“We developed automation finder, creating a portal of opportunities to use automation capabilities or virtual assistant capabilities. Using the technology we created a scoring system — how many times a certain action happened and what systems were used.
“Or, take the customer journey: we can have a look at what journey a customer has been on, so what channels have they used. They’ve phoned up the contact centre, they’ve been on the website, they’ve sent an email, etcetera. Then the software can tie this together, calculating the probability of the customer taking a certain action next, which, armed with data, you can pre-empt.”