Hi
Here is another one, I got couple csv files customers.csv, support.csv, activities.csv.. they are huge files with > 50k records each. Currently they are uploaded to Sql db and using MCP to pull the data. The flow is
- user query goes to LLM via agent (using Google ADK)
- LLM converts that to Sql
- Sql is executed to pull relevant data from the DB
- Agent returns the result back to the user via LLM
All working well. Wondering how can I leverage RAG here and what would be the benefit ? Is RAG primarily for unstructured data ?
Thanks