A local visual RAG challenge for huge construction PDFs

A local system for construction documents needs more than basic chat with a PDF. The first real test file is an 828-page PDF of about 1 GB, with contract text, , schedules, complex tables, and construction drawings mixed together.

Some pages use a large drawing format around 36 by 48 inches, with text near diagrams, callouts, detail tags, and trade-specific sheets where layout matters. The goal is to keep table structure, extract contract language, understand drawing context at a basic level, and answer plain-language questions with .

Accuracy matters more than speed. The hard choices are the , the , which maintained tools to rely on, and which parts should be built directly with ChatGPT, Codex, or Claude instead of premade tools.

Key points

  • The test document is an 828-page construction PDF of about 1 GB.
  • The file combines contract text, , schedules, tables, and drawings.
  • Large-format drawing pages include diagrams, callouts, detail tags, and trade-specific sheets.
  • The target system should preserve tables, extract contract language, read basic drawing context, and cite pages.
  • Accuracy is more important than speed, and tool-vs-custom-build choices are a central concern.
Read original