Retrieval-augmented generation (RAG) is an important tool for LLMs to fetch relevant context from data sources. That's why many platforms offer a way for users to upload a file and have their Agent retrieve context from it based on user queries.
Tasks
[x] Design the front end experience
[ ] Allow users to upload a file from that front end
[ ] Use some smart default chunking strategy
[ ] Create chunk embeddings that can be queried
[ ] Integrate this document as an "Action" for the Agent
[ ] Design a format for annotating this document-search Action
[ ] Return and render the annotation of this RAG feature
Description
Retrieval-augmented generation (RAG) is an important tool for LLMs to fetch relevant context from data sources. That's why many platforms offer a way for users to upload a file and have their Agent retrieve context from it based on user queries.
Tasks