Automated UK Highway Code Q&A and Content Indexing Workflow

This n8n workflow automates the process of extracting, summarizing, embedding, and indexing large UK Highway Code PDFs for advanced search and question-answering functionality. It begins by downloading and segmenting the PDF document into pages, then uses Featherless.ai’s large language model (Kimi-K2) with contextual summaries to generate relevant content chunks. These summaries are embedded using Google Gemini-Embeddings-001, with the vectors stored in a Qdrant vector database. A subworkflow allows querying the indexed data for specific road rules or practical driving questions. Additionally, the workflow supports real-time interactions with a Highway Code expert agent to provide authoritative responses, including a Q&A interface for users seeking specific road rule information or practice questions. Overall, this workflow is ideal for building intelligent, searchable knowledge bases, driving test practice tools, or regulation reference systems leveraging AI and vector search technology.