3. We’re testing an experimental product we've built called GOV.UK Chat, it uses LLMs (in a RAG system) to provide users with quick answers to questions based on 700,000+ pages of GOV.UK 🤖
insidegovuk.blog.gov.uk/2024/11/28/h...

2. We’ve added shortcuts to ‘Popular tasks’ to our browse pages, these are re-generated daily so will reflect seasonal trends in usage 🚀

3. We’ve built an experimental product called GOV.UK Chat which uses Large Language Models (in a RAG system) to get users quick answers to their questions based on the 700,000+ pages of GOV.UK 🤖

Thanks Mark! Appreciate the feedback, we'll be led by the user data for the next iteration but adjustments to header size and spacing are definitely on the cards