Need better understanding of AI Agent and RAG Component
First of all, I'm new to AI Agents in general. In my application, I would like to add a Chat GPT like page where my users can ask questions against my database, essentially query my database with natural language.
So for example a user could ask "What are the top 10 best products sold in the past 3 months?".
I have read the convex documentation about AI Agent and RAG component. But I could not understand how to "direct" toward my tables, or how to "include" my tables in the context. I know about MCP that can "augment" the capabilities of a LLM. With my limited knowledge, I would think something like: "User Prompt" --> Convex Agent --> MCP Tool --> My Convex Database --> (back to Agent) Generating structured response (Not sure if convex has a. built-in way to directly augment the context with tables from DB...) I hope my question make sense and any guidance is appreciated.
I have read the convex documentation about AI Agent and RAG component. But I could not understand how to "direct" toward my tables, or how to "include" my tables in the context. I know about MCP that can "augment" the capabilities of a LLM. With my limited knowledge, I would think something like: "User Prompt" --> Convex Agent --> MCP Tool --> My Convex Database --> (back to Agent) Generating structured response (Not sure if convex has a. built-in way to directly augment the context with tables from DB...) I hope my question make sense and any guidance is appreciated.
1 Reply
Thanks for posting in <#1088161997662724167>.
Reminder: If you have a Convex Pro account, use the Convex Dashboard to file support tickets.
- Provide context: What are you trying to achieve, what is the end-user interaction, what are you seeing? (full error message, command output, etc.)
- Use search.convex.dev to search Docs, Stack, and Discord all at once.
- Additionally, you can post your questions in the Convex Community's <#1228095053885476985> channel to receive a response from AI.
- Avoid tagging staff unless specifically instructed.
Thank you!