# brunolopaiva/local-llm-interface architecture

> Architecture diagram of the GitHub repository brunolopaiva/local-llm-interface, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-11-16). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/brunolopaiva/local-llm-interface
- Repository: https://github.com/brunolopaiva/local-llm-interface

## Overview

ChatLLM is a local document-chat application: users ask questions in a browser, the server retrieves relevant chunks from a knowledge base, and a local Ollama model streams a grounded response back to the chat UI. A separate training page appends new text and rebuilds the in-memory index. The RAG stages and socket handlers are implemented together in server.js, so several nodes share that path. Relationships shown are supported by the server/client excerpts or README; HTML details and other unsampled source were not inspected.

Main entry points: [`server.js`](https://github.com/brunolopaiva/local-llm-interface/blob/main/server.js) (HTTP server), [`public/client.js`](https://github.com/brunolopaiva/local-llm-interface/blob/main/public/client.js) (Chat interface), [`public/treino.js`](https://github.com/brunolopaiva/local-llm-interface/blob/main/public/treino.js) (Training form), [`public/index.html`](https://github.com/brunolopaiva/local-llm-interface/blob/main/public/index.html) (Chat page)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_web["Web experience"]
  node_chat_page["Chat page<br/>[index.html]"]
  node_chat_ui["Chat interface<br/>[client.js]"]
  node_training_page["Training page<br/>[treino.html]"]
  node_training_ui["Training form<br/>[treino.js]"]
end

subgraph group_backend["Application server"]
  node_server["HTTP server<br/>[server.js]"]
  node_socket_handlers["Socket handlers<br/>[server.js]"]
end

subgraph group_rag["Knowledge and RAG"]
  node_knowledge_file["Knowledge text<br/>[data.txt]"]
  node_chunking["Text chunking<br/>[server.js]"]
  node_embeddings["Ollama embeddings<br/>[server.js]"]
  node_vector_store[("Memory vector store<br/>[server.js]")]
  node_retriever["Context retrieval<br/>[server.js]"]
  node_prompt["RAG prompt<br/>[server.js]"]
  node_llm["Local generation<br/>[server.js]"]
end

node_user(("User"))
node_ollama["Ollama service"]

node_user -->|"opens"| node_chat_page
node_chat_page -.->|"loads"| node_chat_ui
node_user -->|"asks"| node_chat_ui
node_chat_ui -->|"sends chat"| node_socket_handlers
node_server -->|"serves static"| node_chat_page
node_server -->|"serves route"| node_training_page
node_training_page -.->|"loads"| node_training_ui
node_user -->|"submits text"| node_training_ui
node_training_ui -->|"sends training"| node_socket_handlers
node_server -->|"hosts sockets"| node_socket_handlers
node_socket_handlers -->|"streams events"| node_chat_ui
node_socket_handlers -->|"sends status"| node_training_ui
node_server -->|"reads at startup"| node_knowledge_file
node_socket_handlers -->|"appends text"| node_knowledge_file
node_server -->|"splits text"| node_chunking
node_server -->|"creates embeddings"| node_embeddings
node_server -->|"builds index"| node_vector_store
node_server -->|"creates retriever"| node_retriever
node_server -->|"configures prompt"| node_prompt
node_server -->|"configures chain"| node_llm
node_socket_handlers -->|"queries chain"| node_retriever
node_socket_handlers -->|"streams generation"| node_llm
node_embeddings -->|"embeds locally"| node_ollama
node_llm -->|"generates locally"| node_ollama
node_socket_handlers -->|"rebuilds RAG"| node_server

click node_chat_page "https://github.com/brunolopaiva/local-llm-interface/blob/main/public/index.html"
click node_chat_ui "https://github.com/brunolopaiva/local-llm-interface/blob/main/public/client.js"
click node_training_page "https://github.com/brunolopaiva/local-llm-interface/blob/main/public/treino.html"
click node_training_ui "https://github.com/brunolopaiva/local-llm-interface/blob/main/public/treino.js"
click node_server "https://github.com/brunolopaiva/local-llm-interface/blob/main/server.js"
click node_socket_handlers "https://github.com/brunolopaiva/local-llm-interface/blob/main/server.js"
click node_knowledge_file "https://github.com/brunolopaiva/local-llm-interface/blob/main/data.txt"
click node_chunking "https://github.com/brunolopaiva/local-llm-interface/blob/main/server.js"
click node_embeddings "https://github.com/brunolopaiva/local-llm-interface/blob/main/server.js"
click node_vector_store "https://github.com/brunolopaiva/local-llm-interface/blob/main/server.js"
click node_retriever "https://github.com/brunolopaiva/local-llm-interface/blob/main/server.js"
click node_prompt "https://github.com/brunolopaiva/local-llm-interface/blob/main/server.js"
click node_llm "https://github.com/brunolopaiva/local-llm-interface/blob/main/server.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_chat_page,node_chat_ui,node_training_page,node_training_ui,node_user toneBlue
class node_server,node_socket_handlers toneAmber
class node_knowledge_file,node_chunking,node_embeddings,node_vector_store,node_retriever,node_prompt,node_llm toneMint
class node_ollama toneIndigo
```

## Components

### Web experience

- [Chat page](https://github.com/brunolopaiva/local-llm-interface/blob/main/public/index.html): `public/index.html`
- [Chat interface](https://github.com/brunolopaiva/local-llm-interface/blob/main/public/client.js): `public/client.js`
- [Training page](https://github.com/brunolopaiva/local-llm-interface/blob/main/public/treino.html): `public/treino.html`
- [Training form](https://github.com/brunolopaiva/local-llm-interface/blob/main/public/treino.js): `public/treino.js`

### Application server

- [HTTP server](https://github.com/brunolopaiva/local-llm-interface/blob/main/server.js): `server.js`
- [Socket handlers](https://github.com/brunolopaiva/local-llm-interface/blob/main/server.js): `server.js`

### Knowledge and RAG

- [Knowledge text](https://github.com/brunolopaiva/local-llm-interface/blob/main/data.txt): `data.txt`
- [Text chunking](https://github.com/brunolopaiva/local-llm-interface/blob/main/server.js): `server.js`
- [Ollama embeddings](https://github.com/brunolopaiva/local-llm-interface/blob/main/server.js): `server.js`
- [Memory vector store](https://github.com/brunolopaiva/local-llm-interface/blob/main/server.js): `server.js`
- [Context retrieval](https://github.com/brunolopaiva/local-llm-interface/blob/main/server.js): `server.js`
- [RAG prompt](https://github.com/brunolopaiva/local-llm-interface/blob/main/server.js): `server.js`
- [Local generation](https://github.com/brunolopaiva/local-llm-interface/blob/main/server.js): `server.js`

### Other components

- **User**
- **Ollama service**

## Connections

- User → Chat page: opens (evidence: `README.md`)
- Chat page → Chat interface: loads
- User → Chat interface: asks (evidence: `README.md`)
- Chat interface → Socket handlers: sends chat (evidence: `public/client.js`)
- HTTP server → Chat page: serves static (evidence: `server.js`)
- HTTP server → Training page: serves route (evidence: `server.js`)
- Training page → Training form: loads
- User → Training form: submits text (evidence: `README.md`)
- Training form → Socket handlers: sends training (evidence: `public/treino.js`)
- HTTP server → Socket handlers: hosts sockets (evidence: `server.js`)
- Socket handlers → Chat interface: streams events (evidence: `server.js`)
- Socket handlers → Training form: sends status (evidence: `server.js`)
- HTTP server → Knowledge text: reads at startup (evidence: `server.js`)
- Socket handlers → Knowledge text: appends text (evidence: `server.js`)
- HTTP server → Text chunking: splits text (evidence: `server.js`)
- HTTP server → Ollama embeddings: creates embeddings (evidence: `server.js`)
- HTTP server → Memory vector store: builds index (evidence: `server.js`)
- HTTP server → Context retrieval: creates retriever (evidence: `server.js`)
- HTTP server → RAG prompt: configures prompt (evidence: `server.js`)
- HTTP server → Local generation: configures chain (evidence: `server.js`)
- Socket handlers → Context retrieval: queries chain (evidence: `server.js`)
- Socket handlers → Local generation: streams generation (evidence: `server.js`)
- Ollama embeddings → Ollama service: embeds locally (evidence: `server.js`)
- Local generation → Ollama service: generates locally (evidence: `server.js`)
- Socket handlers → HTTP server: rebuilds RAG (evidence: `server.js`)
