Company documents and knowledge lived scattered across Drive with no consistent categorisation, no way to see how people, teams and systems connected, and no easy way for the team to just ask a question and get an answer instead of hunting through files.
We built an n8n system that watches a Google Drive folder, and the moment a new document lands, runs it through two AI agents — one categorises and tags it consistently, the other maps it into a knowledge graph of entities and relationships (who reports to whom, what tool a team uses, and so on) — then embeds it into a vector database so a third Librarian agent can answer natural-language questions by searching both the content and the relationship graph.
What we did
Here's how we approached A professional services consultancy's project, step by step.
Auto-detects new documents dropped into a watched Drive folder
Categoriser agent tags every doc consistently (category, subcategory, topics, summary)
Relationship Mapper agent builds a live knowledge graph of people, teams, tools and connections
Documents embedded into a vector database for semantic search
Librarian AI agent answers natural-language questions using both content search and graph lookups
Gap Finder agent flags missing or thin areas in the knowledge base
Results & impact
- Company knowledge organises itself with zero manual filing
- Answers questions like who owns a project or what a team uses, instantly, via the knowledge graph
- An ask-a-question interface replaces manual document searching
- Reusable pattern — works for any team's internal docs, SOPs or client knowledge base
Want something like this for your business?
We build websites, ecommerce stores and AI automation for businesses worldwide — and we'll tell you honestly what you actually need. Book a free 30-minute call and we'll map out your project, costs and timeline.
Book a Free Call