QueryHive 🧠✨

A RAG application built for quering day to day needs

Features:

• Support of openAI & Opensource LLM
• Chuncking/Embedding - cost effective
• cache for LLM - cost effective
• Enabling upload for large documents and data
• CLI options
• Scalability

Audience: Tech/Business (with basic sofware experience)

Design:

Tech Stack:

• UI - ELM
• Backend Services - Rust/Python
• VectorDB - Elastic
• LLM - Opensource/OpenAPI

Setup:

1. Make sure you have docker installed
2. clone the project to local

   git clone https://github.com/Karuturirs/QueryHive.git

3. cd QueryHive
4. Start all applications

   ./start_queryhive.sh

   UI Link Backend Link

5. Stop all applications

   ./stop_queryhive.sh

Other setups

pip install sentence-transformers transformers openai tiktoken nltk

cargo add pyo3 --features "extension-module" cargo add pyo3 --features "auto-initialize"

export RUST_LOG=info # This will enable logs at info level and higher export APP_ENV=local cargo run

TODO List

• UI chat screen to be max
• remove toggle button on screen and make it radio
• work on llm
• why indexing not started on application start
• [ ]