Louisjordaan676/labour-lens

South African labour-law RAG chatbot built as an AI engineering portfolio project. # Labour Lens Labour Lens is a South African labour-law AI assistant designed to help users find and understand information contained in authoritative South African labour-law documents. The project is being developed as an AI engineering portfolio project, with an emphasis on retrieval-augmented generation (RAG), source attribution, software engineering practices, and responsible use of AI for legal-information retrieval. ## Project Status **Current phase:** Development environment setup The application itself has not yet been implemented. The current focus is establishing a clean, reproducible Python development environment and professional project structure. ## Planned Architecture The eventual system is planned to follow this general architecture: ```text Authoritative Labour-Law PDFs ↓ Document Loading ↓ Chunking ↓ Embeddings ↓ Local Vector Store ↓ Retrieval ↓ OpenAI LLM ↓ Answer + Legal Sources ``` The initial implementation will use a local vector store while keeping the application architecture flexible enough to support migration to a managed/cloud vector database in the future. ## Project Structure ```text labour-lens/ │ ├── src/ │ └── labour_lens/ │ └── __init__.py │ ├── data/ │ └── .gitkeep │ ├── tests/ │ └── __init__.py │ ├── .gitignore ├── .env.example ├── README.md └── requirements.txt ``` ## Technology Stack The project is currently being developed with: * Python * LangChain * OpenAI API * Git * GitHub * VS Code Additional technologies will be introduced as the application is developed. ## Development Environment The project uses a Python virtual environment to isolate its dependencies. Python version currently used: ```text Python 3.13.9 ``` ## Important Note Labour Lens is intended to provide information retrieved from authoritative legal sources. It is not intended to replace professional legal advice. ## License License to be determined.