Patrick Daxenbichler - Fotolia

NHS Business Services Authority trials AWS chatbot tech to field customer calls in contact centre

Darren Curry, chief digital officer of the NHS Business Services Authority, shares details of how trialling the use of cloud-based chatbots in its contact centres has boosted staff morale and efficiency

The NHS Business Services Authority (NHSBSA) has said introducing a cloud-based chatbot to its customer contact centres contributed to a reduction in call response times and an increase in staff morale.

The organisation oversees the delivery of primary care dental and prescription services to UK citizens, and is also responsible for providing the NHS workforce with payroll, pensions and other back office support.

Its remit also includes handling applications for the European Health Insurance Card (EHIC) scheme, which provides cardholders with access to free or discounted healthcare services when visiting certain countries in the continent.

EHIC applications and queries are responsible for a sizeable number of the 5 million calls the NHS BSA’s Newcastle-based contact centres receive every year, many of which require a simple yes or no response to be resolved.

“We wanted to enable our contact centre operators to deal with users that required more support,” said Darren Curry, chief digital officer of NHSBSA, during the keynote presentation at the AWS Public Sector Summit in London on Wednesday 9 May.

With this in mind, the organisation started experimenting in November 2017 with rolling out Amazon Web Services (AWS) Connect technology in its contact centres to field some of the more straightforward EHIC queries the service receives.

The technology is billed by the cloud giant as a cloud-based contact centre offering that helps direct calls to the correct operative in a call centre environment, but it can also be integrated with the Amazon Lex chatbot technology to field calls too.

This is the setup NHSBSA opted for and rolled out as a three-week proof of concept near the end of 2017 to capitalise on the seasonal downturn in EHIC calls the service usually receives around this time.

For the bulk of the pilot, the system ran during business hours, but for the final five days, the team set it to run around-the-clock.

“We received just short of 11,000 calls, over the three weeks, [and] 4,300 calls were fully resolved by Lex. That is a 42% success rate,” said Curry.

During the trial period, just 663 of the 11,000 or so calls the service received “dropped out” from Lex, while 5,000 were passed on by the service to human operators to resolve.

Considering the technology setup took just two weeks to build and roll out, with the help of NHSBSA’s cloud broker partner Arcus Global, Curry said the trial is a success story.

Agile and iterative

The NHSBSA takes a “fail-fast” approach to new technology deployments, underpinned by a cloud-first philosophy, Curry told Computer Weekly in a post-keynote interview, so even if it turned out not to work out as planned, it would not have been the end of the world.

“We wanted to make something quickly. If it didn’t work, [we would] let it go, and if it did work, we would move forward,” he said.

The organisation is now in the throes of deciding how best to capitalise on the success of its initial trial, with Curry saying a 24-hour service could be in the offing in future.

Overall, though, the aim of the project was to test whether users would respond well to having their calls dealt with in this way and, in turn, whether introducing it would help reduce the volume of calls its contact centre workers have to deal with.

“We made it very easy for users to drop out to a call centre associate and I don’t think we had many, which was really positive and it proved the hypothesis that if we provided this service people would be willing to use it on some of the types of calls we receive,” he said.

Another added benefit is the positive impact deploying the technology had on the contact centre’s staff. Their feedback played a key role in helping shape the prototype service, and the fact it helped reduce their workload had a positive impact on morale too, added Curry.

“[Our staff] want to be delivering value calls, so it was a really positive experience. They look to resolve calls as quickly as possible, but it enabled them to get a better sense of self-gratification by dealing with those more value-add service queries,” he said.

Read more about cloud and machine learning

Read more on Infrastructure-as-a-Service (IaaS)

Data Center
Data Management