Where next for AI? Understanding business use cases

Exploiting information to create new intelligence

Barry Libenson, global CIO at financial data company Experian, says emerging technologies such as AI and ML are going to be “an absolute game changer”, adding that, in simple terms, ML will allow businesses to create better insights.

Libenson believes digital leaders can incorporate information, such as social media data, with financial data and create new knowledge. 

“There are some really interesting things that we’ve seen,” he says. “Some people are so affluent that they never borrow money and they may not have a strong credit history. But if you look at certain aspects of somebody’s social media profile – and you know what things they like, where they hang out, what they do – you can actually start to build insights based on social media data that are indicative of behaviour.” 

Libenson says another potential interesting application of ML is in the of infrastructure management. “We’ll see certain behaviours on a machine and we can learn from that,” he says. “This isn’t anything new, but the technology’s got a lot better. You can do diagnostics, so you can start to determine when something’s going to go wrong based on certain things that you see.” 

In some cases, says Libenson, a human being can’t necessarily see seeing subtle changes, whereas a machine is much more capable. His team takes data from its DynaTrace logs, which monitor system behaviour, and feeds that information into Splunk. This ML-led approach looks for patterns in data that a human being wouldn’t necessarily detect. 

“Splunk can see it because of its ability to process the information,” says Libenson. “If it sees certain things, it’ll alert somebody in the operation centre and say that the DynaTrace logs and Splunk are reporting some behaviour that’s out of the ordinary, so someone should look into it and see what’s going on.” 

Libenson says this knowledge positively affects business performance. “Something like that could prevent a system outage or a failover to another device,” he adds. “There’s lots of interesting applications for ML right now.”

Using the cloud as a means to explore emerging technology 

Guinness World Records (GWR) IT director Rob Howe has supported a business transformation at his organisation through the use of digital technology. The firm has evolved during Howe’s six years with the organisation from just being a publishing house that produces its iconic book of records, to a creative consultancy that works alongside brands on marketing campaigns. 

Howe has led a staged approach to digital transformation that has included the implementation of a records management platform from SDL, a digital asset management system from Asset Bank, and Salesforce CRM technology. As a final stage of the transformation process, GWR chose Ensono to manage the migration of its business-critical IT architecture to an Amazon Web Services (AWS) platform. 

Howe aims to use the cloud as a platform for further innovation. He says the next step is to convert GWR’s application programming interface (API) layer to microservices. Then Howe and his team will think about what type of data should be pushed out to edge locations. Finally, he might consider how the firm can take advantage of ML, although he recognises that stage is some way in advance. 

“It’s an idea, but it needs more thought. It’s a potential solution to one of the challenges we face as an organisation, but it’s been parked until we’re stable and more confident in the platform. It’s still early days for us – our infrastructure on AWS went live at the end of September,” says Howe. 

“We’re going to be looking at upgrading our asset platform as new versions of that service are launched. We could also look at whether we could use ML to help process some of the record applications more effectively and in an automated manner. Now we’ve done the move, we need to stabilise, show the areas where we’ve added value and engage the rest of the company in terms of new services.” 

Embracing AI to free up valuable human resources 

Darren Curry, chief digital officer at NHS Business Services Authority (NHSBSA), is also leading a digital transformation programme. The journey started in 2015, when NHSBSA started to think about how it would digitise the paper-based records of the maternity exemption service it manages. 

“Back when we started thinking about transformation, AI wasn’t on our agenda,” says Curry, who suggests AI has “huge potential” in the NHS.

NHSBSA has already embraced a cloud-first strategy for hosting. Alongside this shift to on-demand IT, Curry is keen to ensure emerging technology, such as AI, is exploited – and that work is already underway. 

“We get about 4.7 million contacts from customers every year in our 600-seat contact centre in Newcastle. We receive pension calls, requests from GPs, and other health-related enquiries. We recently introduced AI using Amazon Alexa in the contact centre for calls relating to the European Health Insurance Card (EHIC),” says Curry.

Curry says NHSBSA implemented the technology in two weeks – that was from concept to going live. After four weeks, his team ramped the service up to 24/7. The technology has helped support a 45% reduction in the amount of calls to contact centre operators. The aim has been to remove non-complex calls from getting through to an operator. 

“AI will help us deal with questions like, ‘Can I use an EHIC in Australia?’. We got a huge reduction in calls by applying the technology – and the service is now running live for EHIC. We also intend to consider AI for some of the other streams of work and to roll Amazon Alexa out across our call centre operations,” says Curry. 

Investigating how cognitive capabilities can boost operational activities

Ian Cohen, CIO at transport specialist Addison Lee, says effective use of emerging technology is related to safe use of customer data. He says there are some obligations that an organisation must respect when a customer allows them to have their data. 

“We’ve gone through a decade of people trading their personal information for free Wi-Fi or to play a game on Facebook,” he says. “Hopefully, people in the future will be more circumspect and understand the true value of their data.” 

As part of this forward-looking process, Addison Lee is investigating how autonomous vehicles might be developed and used in global cities. Cohen says his firm holds more than 30 years of data around how an organisation allocates and dispatches cars to fulfil customer requests around London. 

“We complete around 25,000 journeys per day and over 90% are auto-allocated based on driver availability, traffic conditions and traffic services using a set of algorithms. Where that goes next, in terms of intuitive and interpretative ML, is interesting,” he says, before placing developments at his firm within a broader, industry-wide context. 

“We need to understand what happens when you connect things up and how you learn and derive insight from those environments, and then how you take the learning into a programmatic state. True AI is a long way off – people are still taking their first steps with ML. Cognitive, contextual and situationally aware AI is for the future.”

Using data to create value for stakeholders and customers 

Rob McLaughlin, head of digital decisioning and analytics at Sky, says the use of emerging technology is best seen as an extremely complex area where a lot of things are happening simultaneously. He says many executives talk about using AI, yet their firms still haven’t addressed some of the basics. 

“ML is best understood as a set of statistical techniques which you can use for anything – it doesn’t have to be used to act; it can be used for analysing data sets,” says McLaughlin. “AI is more about making some form of decision; it’s more applied. AI is almost always associated with leveraging ML.” 

McLaughlin says human-based rules are useful when it comes to explaining things to his firm’s stakeholders about how decisions are made – so, if someone likes sport, the services team might recommend football-based products. However, the relationships suggested by the black-box technology of AI can be trickier to pick up and explain to stakeholders. 

“We’re creating capability – and it then needs to be adopted by our business, such as plugging the application programming interface into our home page, contact centre systems or the mobile app. That adoption requires strong relationships with people around our business and managing those can be a complex process,” says McLaughlin. 

His team ensures their activities match stakeholder requirements. Their efforts must work towards at least one of three business objectives: up-sell and cross-sell, which is about recommending products; in-life engagements, ensuring customers use more of the products they have; and service messaging, which is about dealing with clients in an efficient way. 

“Data has to be a positive thing – it has to help keep your customers happy,” says McLaughlin, referring to his team’s work around personalisation and emerging technology. “There’s no future for data that doesn’t help build the relationship between consumers and businesses. A genuine value exchange must be the priority for all customer-facing organisations.”