# ayushshahir/youtube-rag-chatbot architecture

> Architecture diagram of the GitHub repository ayushshahir/youtube-rag-chatbot, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-10-06). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/ayushshahir/youtube-rag-chatbot
- Repository: https://github.com/ayushshahir/youtube-rag-chatbot

## Overview

This is a YouTube transcript RAG chatbot. A user loads a video, the backend retrieves and chunks its transcript, indexes it for semantic search, and answers questions with Gemini; the interface presents timestamp-linked answers and supports conversation memory. The README also documents Hindi and Marathi translation. The graph reflects the excerpts and README; unsampled frontend and backend details limit some interaction links.

Main entry points: [`backend/app.py`](https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py) (Flask API), [`static/script.js`](https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/static/script.js) (Chat interface), [`templates/index.html`](https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/templates/index.html) (Chat page)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_web["Chat interface"]
  node_ui["Chat interface<br/>[script.js]"]
  node_page["Chat page<br/>[index.html]"]
  node_translation["Answer translation<br/>[script.js]"]
end

subgraph group_ingestion["Video ingestion"]
  node_api["Flask API<br/>[app.py]"]
  node_transcript["Transcript fetch<br/>[app.py]"]
  node_chunking["Timestamped chunks<br/>[app.py]"]
  node_embeddings["Text embeddings<br/>[app.py]"]
  node_vectors[("Vector store<br/>[app.py]")]
end

subgraph group_answering["Retrieval and answers"]
  node_retrieval["Semantic retrieval<br/>[app.py]"]
  node_rag["RAG chain<br/>[app.py]"]
end

subgraph group_state["Conversation and links"]
  node_memory[("Conversation memory<br/>[app.py]")]
  node_timestamps["Timestamp links<br/>[app.py]"]
end

node_user(("Video viewer"))
node_youtube["YouTube transcript service"]
node_gemini["Gemini model"]
node_youtube_player["YouTube player"]

node_user -->|"uses"| node_ui
node_api -->|"renders"| node_page
node_ui -->|"processes video"| node_api
node_api -->|"fetches"| node_transcript
node_transcript -->|"requests transcript"| node_youtube
node_api -->|"chunks transcript"| node_chunking
node_api -->|"creates store"| node_vectors
node_vectors -->|"uses"| node_embeddings
node_api -->|"configures"| node_retrieval
node_rag -->|"retrieves context"| node_retrieval
node_rag -->|"generates answer"| node_gemini
node_api -->|"invokes"| node_rag
node_api -->|"tracks turns"| node_memory
node_api -->|"builds sources"| node_timestamps
node_ui -->|"translates answers"| node_translation
node_ui -->|"seeks timestamps"| node_youtube_player

click node_ui "https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/static/script.js"
click node_page "https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/templates/index.html"
click node_api "https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py"
click node_transcript "https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py"
click node_chunking "https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py"
click node_embeddings "https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py"
click node_vectors "https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py"
click node_retrieval "https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py"
click node_rag "https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py"
click node_memory "https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py"
click node_timestamps "https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py"
click node_translation "https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/static/script.js"

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_ui,node_page,node_translation,node_user toneBlue
class node_api,node_transcript,node_chunking,node_embeddings,node_vectors toneAmber
class node_retrieval,node_rag toneMint
class node_memory,node_timestamps toneRose
class node_youtube,node_gemini,node_youtube_player toneIndigo
```

## Components

### Chat interface

- [Chat interface](https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/static/script.js): `static/script.js`
- [Chat page](https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/templates/index.html): `templates/index.html`
- [Answer translation](https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/static/script.js): `static/script.js`

### Video ingestion

- [Flask API](https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py): `backend/app.py`
- [Transcript fetch](https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py): `backend/app.py`
- [Timestamped chunks](https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py): `backend/app.py`
- [Text embeddings](https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py): `backend/app.py`
- [Vector store](https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py): `backend/app.py`

### Retrieval and answers

- [Semantic retrieval](https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py): `backend/app.py`
- [RAG chain](https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py): `backend/app.py`

### Conversation and links

- [Conversation memory](https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py): `backend/app.py`
- [Timestamp links](https://github.com/ayushshahir/youtube-rag-chatbot/blob/main/backend/app.py): `backend/app.py`

### Other components

- **Video viewer**
- **YouTube transcript service**
- **Gemini model**
- **YouTube player**

## Connections

- Video viewer → Chat interface: uses
- Flask API → Chat page: renders (evidence: `backend/app.py`)
- Chat interface → Flask API: processes video (evidence: `static/script.js`)
- Flask API → Transcript fetch: fetches (evidence: `backend/app.py`)
- Transcript fetch → YouTube transcript service: requests transcript (evidence: `backend/app.py`)
- Flask API → Timestamped chunks: chunks transcript (evidence: `backend/app.py`)
- Flask API → Vector store: creates store (evidence: `backend/app.py`)
- Vector store → Text embeddings: uses (evidence: `backend/app.py`)
- Flask API → Semantic retrieval: configures (evidence: `backend/app.py`)
- RAG chain → Semantic retrieval: retrieves context (evidence: `backend/app.py`)
- RAG chain → Gemini model: generates answer (evidence: `backend/app.py`)
- Flask API → RAG chain: invokes (evidence: `backend/app.py`)
- Flask API → Conversation memory: tracks turns (evidence: `backend/app.py`)
- Flask API → Timestamp links: builds sources (evidence: `backend/app.py`)
- Chat interface → Answer translation: translates answers (evidence: `README.md`)
- Chat interface → YouTube player: seeks timestamps (evidence: `static/script.js`)
