United KingdomPublic sector2025 – ongoingNational Highways
Language-understanding webchat for road enquiries
National Highways deployed a language-understanding chatbot handling approximately 100 engaged sessions monthly
Problem
Customers relied on existing contact channels to ask about traffic and road closures. National Highways wanted immediate self-service answers for routine enquiries.
Approach
National Highways worked with delivery partner Kainos-Arup on webchat using Microsoft's language-understanding model. Staff added examples of customer phrasing so the model could select relevant topics. Responses followed approved scripts and public information.
Results
Takeaway
The learned language model interpreted requests while staff controlled the answers. The disclosed session count measured use, without establishing a reduction in waiting times.
Categories
- Industry
- Public sector
- Business function
- Customer service
- Innovation type
- Process innovation
- AI technology
- Conversational and voice AI
- Adoption
- Partner co-development
- Evidence
- Company's own claim
Sources
Source summaries
The transparency record documented a production chatbot using trained language understanding with pre-programmed answers. National Highways reported approximately 100 engaged sessions monthly and retained other contact channels.
Full details are in the original sources.