n8n PDF-to-LLM Workflow Build

I need an n8n flow that ingests research-grade PDFs packed with headings, sub-headings, tables, diagrams and dense body text, then automatically extracts the salient data, curates it, and produces clear, concise summaries with the help of an LLM.

Core actions in the workflow
• Data extraction – reliably pull text, tables and visuals from every page.
• Content curation – organise those fragments into a clean, logical structure ready for the model.
• Content summarisation – call the LLM and return a well-formatted synopsis.

Deliverables
1. A reusable n8n workflow (.json) covering the full pipeline from PDF upload to summary output.
2. Brief setup notes so I can drop in my own API keys and run it locally or on n8n Cloud.
3. One example run using a sample PDF that proves extraction, curation and summarisation all work end-to-end.