CROP

RAG Pipelines

Retrieval-augmented generation pipelines for parts lookup

RAG Pipelines

Retrieval-augmented generation pipelines that combine vector search with LLM generation for intelligent parts lookup.

Repository

CROP-RAG-Pipelines

Overview

  • Query understanding and intent classification
  • Hybrid search: Typesense (keyword) + Weaviate (semantic)
  • LLM-augmented response generation
  • Parts recommendation and cross-reference

Stack

  • Python, LangChain
  • Weaviate for vector search
  • Integration with PDF Parser and Embedding Service

On this page