The Driver and Vehicle Licensing Agency has used artificial intelligence technology to make a big reduction in the average length of the calls made by 900,000 citizens each month to the organisation’s contact centres.
The agency has deployed an interactive voice response system based on natural language processing (IVR and NLP), recently released transparency documents reveal. The AI tool is used to field citizens’ calls to DVLA helplines addressing queries related to general licensing issues, medical considerations, vehicle management and personal registration, and problems with using digital services, including viewing licences online.
But, alongside these limitations, the department has “diversified our entry routeways including an increased focus on our other life-chances schemes such as Movement to Work, in addition to apprenticeship opportunities”, Western said, in response to a written parliamentary question from Conservative MP Peter Bedford.
When callers ring the agency’s telephone support lines, the IVR , which is based on AI technology from Google, “asks a customer to ‘tell us’ what their enquiry is about using their voice, [and] the NLP engine uses this intent to route to pre-programmed options within the IVR”, according to an algorithmic transparency record newly published on GOV.UK.
The document adds that the options to which citizens may be directed include: “messages that will attempt to answer the customer’s enquiry; options to receive an SMS which contains a link to a DVLA service or knowledge article on GOV.UK; [and] routing customers to the appropriate advisor”.
The algorithmic record says that, “if a customer has their enquiry answered by a message in the IVR, then they may hang up the call without the need to speak to an advisor”.
As a result of which, since implementation of the NLP system, the DVLA claims that the average time callers spend navigating the automated options given to them at the start of their call has been cut in half: from about three minutes to 90 seconds.
Before installing the AI-powered system, the DVLA used a touchtone platform which required callers to listen to various choices and then press a button to route their call accordingly.
“To allow us to understand the enquiry type, these menus often included four or five options and were three or four levels deep,” the algorithmic record states. “This created a clunky customer journey and did not achieve a high level of routing accuracy.”
The AI tool has also helped automated the transfer to the appropriate adviser of 20,000 calls each month, according to the transparency document.
“NLP IVR has also provided the business with more granular and accurate call descriptors enabling the business to make more informed decisions around training and coaching requirements and routing accuracy,” the record adds.
The tool is based the Google AI Dialogflow software and has been deployed via the Storm contact centre system offered by Content Guru – the firm that provides the DVLA with the platform that underpins its helplines.
In an interview earlier this year, DVLA's chief executive Tim Moss told CSW that the agency is “pushing the boundaries of AI". “When you look at how they’re using AI for digital development and some of the agentic AI stuff, to me this is almost black magic,” he said.
Moss also said that AI can help with the agency’s task of finding 15% efficiency savings over the Spending Review 2025 period, but its "primary focus has to be around the customer experience”.
“For me, it’s not about saying: how can we replace people or do things cheaper? It’s about saying: how can we use AI to deliver a better customer experience? That’s what we’re here to do. Once we get to the level of customer experience we want, we can then look at how we make this more efficient, more effective.”