AI Chat Module

Chat with AI based on your vault's notes. Uses RAG (Retrieval-Augmented Generation) to automatically find relevant notes and provide them as context.

Usage

1. Open the Command Palette with Cmd+Shift+P
2. Select "AI Chat"
3. Enter your question
4. Click Enter or "Send"

Features

RAG (Retrieval-Augmented Generation)

When you enter a question:

1. Converts the question into an embedding vector
2. Searches for semantically similar notes in the Vault
3. Passes relevant notes as context to the LLM
4. LLM generates a response based on the context

Visual RAG - Display Sources

Referenced notes are displayed as links under the AI response. Click to go to that specific note.

Local Processing

All processing occurs locally. Data is not sent externally.

Vault Indexing

You must index your Vault before using AI Chat:

1. Settings → Naidis → AI
2. Click "Index Now"
3. Wait for indexing to complete (time varies by number of notes)

Re-index after adding new notes.

Requirements

Ollama (Recommended)

curl -fsSL https://ollama.com/install.sh | sh
ollama pull llama3.2

Changing Models

You can change the model to be used in Settings → Naidis → AI:

• llama3.2 (Default, lightweight)
• llama3.1 (More powerful)
• mistral
• Other Ollama-supported models

Tips

• More specific questions lead to more accurate answers
• Questions like "Summarize meeting minutes related to Project X" are effective
• The first question after indexing may take time due to model loading