Pinterest's RAG-based table selection system operates by generating a vector index of table summaries, transforming user questions into embeddings, conducting a similarity search to infer the top N suitable tables, using an LLM to select the top K most suitable tables, and finally creating a text-to-SQL prompt to generate the response.

Authors

• Person: Not available Organization: Evidently AI
  10 RAG examples and use cases from real companies - Evidently AI

Sources

• 10 RAG examples and use cases from real companies - Evidently AI www.evidentlyai.com Evidently AI via serper

Referenced by nodes (1)

• RAG concept