Local Context Engine

A privacy-focused, local RAG (Retrieval-Augmented Generation) application that allows you to chat with your PDF documents using Llama 3.2 and Ollama.

Features

• 100% Local: No data leaves your machine.
• Fast Inference: Uses the optimized Llama 3.2 SLM.
• Modern UI: Built with Streamlit, featuring a dark mode, glassmorphism design, and responsive feedback mechanisms.
• Feedback Loop: Built-in 👍/👎 system to log hallucinations and improve performance.

Prerequisites

1. Python 3.8+
2. Ollama: You must have Ollama installed and running.
   • Download Ollama
   • Pull the model: ollama pull llama3.2

Installation

1. Clone the repository:

   git clone <repository-url>
   cd "Local Context Engine"

2. Create a virtual environment (optional but recommended):

   python -m venv .venv
   source .venv/bin/activate  # On Windows: .venv\Scripts\activate

3. Install dependencies:

   pip install streamlit langchain langchain-community chromadb

Usage

1. Start the application:

   streamlit run app.py

2. Upload Documents:

   • Open the sidebar.
   • Drag and drop your PDF files.
   • Click Process Documents.

3. Chat:

   • Ask questions about your documents in the chat input.
   • View the latency metrics for each response.
   • Use the 👍/👎 buttons to provide feedback (logged to logs.json).

Project Structure

• app.py: Main Streamlit application and UI logic.
• rag_engine.py: Core RAG logic (loading, splitting, embedding, retrieving).
• logs.json: Stores user feedback and performance metrics.
• SHIP_REPORT.md: Detailed analysis of the build process and model performance.