How nisaaragharia/advanced_rag works
Advanced_RAG is a collection of LangChain notebooks teaching RAG techniques, rather than a single integrated runtime application. The documented basic workflow runs from a user query to a generated response; the remaining notebooks cover distinct retrieval and agentic approaches. Source excerpts were unavailable, so relationships beyond the README’s stated basic flow are not inferred.
Main entry points: 01_Introduction_To_RAG.ipynb Basic RAG
Components 12
RAG Foundations
- Basic RAG
01_Introduction_To_RAG.ipynb - Query Transformations
02_Query_Transformations.ipynb - Datasource Routing
03_Routing_To_Datasources.ipynb - Vector Indexing
04_Indexing_To_VectorDBs.ipynb
Retrieval Techniques
- Retrieval Mechanisms
05_Retrieval_Mechanisms.ipynb
Reflective Agents
- Self-Reflection RAG
06_Self_Reflection_Rag.ipynb - Agentic RAG
07_Agentic_Rag.ipynb - Adaptive Agentic RAG
08_Adaptive_Rag_Agent.ipynb - Corrective Agentic RAG
09_Corrective_RAG_Agent.ipynb
Local Agent
- Local Llama Agent
10_LLAMA_3_Rag_Agent_Local.ipynb
Other components
- User
- Generated Response
Connections 2
- User → Basic RAG: submits query
README.md - Basic RAG → Generated Response: generates response
README.md