# shruti-06/cuisine-knowledge-rag-system architecture

> Architecture diagram of the GitHub repository shruti-06/cuisine-knowledge-rag-system, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-11-19). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/shruti-06/cuisine-knowledge-rag-system
- Repository: https://github.com/shruti-06/cuisine-knowledge-rag-system

## Overview

This repository is a notebook-driven South Asian cuisine RAG application. A user submits a free-text question or JSON batch through Gradio; questions are retrieved against the prepared corpus, reranked, and answered by a grounded language model, with retrieved context returned to the user. The notebook also documents corpus preparation, chunking, embedding, and benchmark evaluation. Relationships follow the README’s documented pipeline; source code was not excerpted, so implementation-level call ownership cannot be verified.

Main entry points: [`Notebook/Codebase.ipynb`](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb) (Dense retrieval), [`Indexes/bge_chunks.pkl`](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Indexes/bge_chunks.pkl) (Chunk store), [`Data/Background_Corpus.json`](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Data/Background_Corpus.json) (Cleaned corpus), [`Indexes/bge_index.faiss`](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Indexes/bge_index.faiss) (FAISS vector index)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_corpus["Corpus preparation"]
  node_wiki(("Wikipedia sources"))
  node_wikibooks(("Wikibooks sources"))
  node_blog(("Food blog sources"))
  node_wiki_pipeline["Wikipedia collection<br/>[Codebase.ipynb]"]
  node_wikibooks_pipeline["Wikibooks collection<br/>[Codebase.ipynb]"]
  node_blog_pipeline["Blog collection<br/>[Codebase.ipynb]"]
  node_corpus_store[("Cleaned corpus")]
  node_chunking["Adaptive chunking<br/>[Codebase.ipynb]"]
end

subgraph group_retrieval["Retrieval"]
  node_chunk_store[("Chunk store<br/>[bge_chunks.pkl]")]
  node_embedding["BGE embeddings<br/>[Codebase.ipynb]"]
  node_faiss[("FAISS vector index<br/>[bge_index.faiss]")]
  node_dense["Dense retrieval<br/>[Codebase.ipynb]"]
  node_bm25["BM25 retrieval<br/>[Codebase.ipynb]"]
  node_fusion["Hybrid fusion<br/>[Codebase.ipynb]"]
  node_reranker["Cross-encoder reranking<br/>[Codebase.ipynb]"]
end

subgraph group_answering["Answering and evaluation"]
  node_generation["Grounded generation<br/>[Codebase.ipynb]"]
  node_benchmark[("Benchmark dataset")]
  node_evaluation["Retrieval and answer evaluation<br/>[Codebase.ipynb]"]
end

subgraph group_interface["User interface"]
  node_gradio["Gradio interface<br/>[Codebase.ipynb]"]
  node_batch["JSON batch route<br/>[Codebase.ipynb]"]
end

node_user(("User"))

node_wiki -->|"provides pages"| node_wiki_pipeline
node_wikibooks -->|"provides pages"| node_wikibooks_pipeline
node_blog -->|"provides posts"| node_blog_pipeline
node_wiki_pipeline -->|"cleans and collects"| node_corpus_store
node_wikibooks_pipeline -->|"cleans and collects"| node_corpus_store
node_blog_pipeline -->|"cleans and collects"| node_corpus_store
node_corpus_store -->|"chunks documents"| node_chunking
node_chunking -->|"stores chunks"| node_chunk_store
node_chunk_store -->|"embeds chunks"| node_embedding
node_embedding -->|"builds index"| node_faiss
node_faiss -->|"supplies vectors"| node_dense
node_chunk_store -->|"provides chunks"| node_dense
node_chunk_store -->|"provides chunks"| node_bm25
node_dense -->|"contributes scores"| node_fusion
node_bm25 -->|"contributes scores"| node_fusion
node_fusion -->|"ranks candidates"| node_reranker
node_reranker -->|"provides context"| node_generation
node_benchmark -->|"supplies queries"| node_evaluation
node_dense -->|"evaluates retrieval"| node_evaluation
node_generation -->|"evaluates answers"| node_evaluation
node_user -->|"submits questions"| node_gradio
node_user -->|"uploads queries"| node_batch
node_batch -->|"routes batch input"| node_gradio
node_gradio -->|"retrieves context"| node_dense
node_gradio -->|"requests answers"| node_generation
node_generation -->|"returns grounded answer"| node_gradio

click node_wiki_pipeline "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb"
click node_wikibooks_pipeline "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb"
click node_blog_pipeline "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb"
click node_corpus_store "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Data/Background_Corpus.json"
click node_chunking "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb"
click node_chunk_store "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Indexes/bge_chunks.pkl"
click node_embedding "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb"
click node_faiss "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Indexes/bge_index.faiss"
click node_dense "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb"
click node_bm25 "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb"
click node_fusion "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb"
click node_reranker "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb"
click node_generation "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb"
click node_benchmark "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Data/RAG_Benchmark_Dataset.json"
click node_evaluation "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb"
click node_gradio "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb"
click node_batch "https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_wiki,node_wikibooks,node_blog,node_wiki_pipeline,node_wikibooks_pipeline,node_blog_pipeline,node_corpus_store,node_chunking,node_user toneBlue
class node_chunk_store,node_embedding,node_faiss,node_dense,node_bm25,node_fusion,node_reranker toneAmber
class node_generation,node_benchmark,node_evaluation toneMint
class node_gradio,node_batch toneRose
```

## Components

### Corpus preparation

- **Wikipedia sources**
- **Wikibooks sources**
- **Food blog sources**
- [Wikipedia collection](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb): `Notebook/Codebase.ipynb`
- [Wikibooks collection](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb): `Notebook/Codebase.ipynb`
- [Blog collection](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb): `Notebook/Codebase.ipynb`
- [Cleaned corpus](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Data/Background_Corpus.json): `Data/Background_Corpus.json`
- [Adaptive chunking](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb): `Notebook/Codebase.ipynb`

### Retrieval

- [Chunk store](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Indexes/bge_chunks.pkl): `Indexes/bge_chunks.pkl`
- [BGE embeddings](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb): `Notebook/Codebase.ipynb`
- [FAISS vector index](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Indexes/bge_index.faiss): `Indexes/bge_index.faiss`
- [Dense retrieval](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb): `Notebook/Codebase.ipynb`
- [BM25 retrieval](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb): `Notebook/Codebase.ipynb`
- [Hybrid fusion](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb): `Notebook/Codebase.ipynb`
- [Cross-encoder reranking](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb): `Notebook/Codebase.ipynb`

### Answering and evaluation

- [Grounded generation](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb): `Notebook/Codebase.ipynb`
- [Benchmark dataset](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Data/RAG_Benchmark_Dataset.json): `Data/RAG_Benchmark_Dataset.json`
- [Retrieval and answer evaluation](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb): `Notebook/Codebase.ipynb`

### User interface

- [Gradio interface](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb): `Notebook/Codebase.ipynb`
- [JSON batch route](https://github.com/shruti-06/cuisine-knowledge-rag-system/blob/main/Notebook/Codebase.ipynb): `Notebook/Codebase.ipynb`

### Other components

- **User**

## Connections

- Wikipedia sources → Wikipedia collection: provides pages (evidence: `README.md`)
- Wikibooks sources → Wikibooks collection: provides pages (evidence: `README.md`)
- Food blog sources → Blog collection: provides posts (evidence: `README.md`)
- Wikipedia collection → Cleaned corpus: cleans and collects (evidence: `README.md`)
- Wikibooks collection → Cleaned corpus: cleans and collects (evidence: `README.md`)
- Blog collection → Cleaned corpus: cleans and collects (evidence: `README.md`)
- Cleaned corpus → Adaptive chunking: chunks documents (evidence: `README.md`)
- Adaptive chunking → Chunk store: stores chunks (evidence: `README.md`)
- Chunk store → BGE embeddings: embeds chunks (evidence: `README.md`)
- BGE embeddings → FAISS vector index: builds index (evidence: `README.md`)
- FAISS vector index → Dense retrieval: supplies vectors (evidence: `README.md`)
- Chunk store → Dense retrieval: provides chunks (evidence: `README.md`)
- Chunk store → BM25 retrieval: provides chunks (evidence: `README.md`)
- Dense retrieval → Hybrid fusion: contributes scores (evidence: `README.md`)
- BM25 retrieval → Hybrid fusion: contributes scores (evidence: `README.md`)
- Hybrid fusion → Cross-encoder reranking: ranks candidates (evidence: `README.md`)
- Cross-encoder reranking → Grounded generation: provides context (evidence: `README.md`)
- Benchmark dataset → Retrieval and answer evaluation: supplies queries (evidence: `README.md`)
- Dense retrieval → Retrieval and answer evaluation: evaluates retrieval (evidence: `README.md`)
- Grounded generation → Retrieval and answer evaluation: evaluates answers (evidence: `README.md`)
- User → Gradio interface: submits questions (evidence: `README.md`)
- User → JSON batch route: uploads queries (evidence: `README.md`)
- JSON batch route → Gradio interface: routes batch input (evidence: `README.md`)
- Gradio interface → Dense retrieval: retrieves context (evidence: `README.md`)
- Gradio interface → Grounded generation: requests answers (evidence: `README.md`)
- Grounded generation → Gradio interface: returns grounded answer (evidence: `README.md`)
