Better context starts before the prompt.
Prepare a knowledge collection with meaningful chunks, source lineage, permission boundaries, and evaluation grounded in real questions.
Read the guideImport Notes · Topic
Documents
Turn files into coherent, traceable sections. Consider reading order, tables, revision identity, and extraction quality before routing material into search or an AI knowledge collection.
Give your documents a second life.
Build a document import that preserves headings, handles uncertain extraction, and connects useful text to its original source and revision.
Read the guideConnect the ideas to a workflow
Use these readings to define a small, testable handoff: an approved source, explicit destination fields, and a correction path for ambiguous values. Keep the original representation visible while checking the proposed transformation.
Explore the import categories for source-specific decisions or follow the local developer quickstart to see a complete synthetic mapping and validation result.