hardDesignScenarioMidSeniorStaff
Your corpus is scanned PDFs, spreadsheets and slide decks. How do you ingest it?
6 min readUpdated 2026-08-15
Answer
Short answer
Route each format to a parser that respects its structure, treat OCR output as text with a confidence score rather than as truth, convert tables into rows that carry their headers, and record a section path and stable anchor per chunk so every retrieved passage can be cited back to a place in the original.
Was this answer helpful?
Likely follow-ups
- A scanned page comes back as gibberish. How does that get caught before it reaches the index?
- A spreadsheet has merged cells and three tables on one sheet. What do you do with it?
- How do you handle a diagram on a slide whose meaning is entirely visual?
- Which of these formats would you refuse to support in the first release, and how would you argue it?
Related questions
- How do you choose a chunking strategy for a document corpus?mediumAlso on ingestion and rag4 min
- What metadata do you store alongside each chunk, and what is each field for?mediumAlso on ingestion and rag4 min
- Your RAG assistant cites the wrong document, and sounds completely sure of itself. How do you find which stage failed?hardAlso on rag5 min
- You are building a support assistant over a knowledge base that changes every week. RAG or fine-tuning, and how do you decide?mediumAlso on rag5 min