Aktu-Rag: Local Academic AI Assistant

A production-grade Local RAG CLI application that indexes and answers questions from notes, PDFs, and study materials completely offline.

Features

• Fully local and offline inference using Ollama
• Intelligent text chunking and subject detection
• OCR support for scanned notes using Tesseract
• Persistent vector storage via ChromaDB
• CLI commands with Typer

Requirements

• Python 3.12+
• tesseract for OCR (brew install tesseract on macOS, apt-get install tesseract-ocr on Ubuntu)
• poppler for PDF manipulation (brew install poppler on macOS, apt-get install poppler-utils on Ubuntu)
• Ollama installed locally

Installation

1. Setup the environment:

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

2. Make sure Ollama is running, and pull the default model:

ollama pull llama3

3. Ensure aktu-rag is executable:

chmod +x aktu-rag

Usage

Ingesting Notes

Place your PDFs and text files in the notes/ directory, then run:

./aktu-rag ingest ./notes

Asking Questions

./aktu-rag ask "Explain Banker's Algorithm"

View Indexed Documents

./aktu-rag list-docs

Index Statistics

./aktu-rag stats

Clear and Rebuild

./aktu-rag rebuild-index

Docker Support

You can run the app alongside Ollama using Docker Compose:

docker-compose up -d
docker-compose exec rag-app bash
# Then run commands inside the container
./aktu-rag ingest ./notes