RAG design for CV search shows where cost and quality collide

A CV search tool for recruiters is being designed around three steps: , , and LLM processing. The starting size is a few thousand CVs. The hardest part is the parsing layer, where messy CV files must be turned into useful search data.

Fixed-size chunks can break important context. Splitting by sections, such as work history or education, can be cleaner, but real CV formats vary a lot. Another option is to use an LLM during to extract structured fields before search happens.

That could improve later searches, but it may become expensive or error-prone as the grows. If the metadata is wrong at the start, the later search and LLM output can become unreliable. Reliable PDF is also a core requirement because real-world CVs are often messy.

Key points

  • The planned flow is , , then LLM processing.
  • Fixed-size chunks can lose context from a CV.
  • Section-based parsing can be cleaner, but CV layouts are inconsistent.
  • Using an LLM during may reduce later search cost, but it adds upfront cost and scaling risk.
  • Bad metadata can weaken every later step in the search pipeline.
Read original