foodie

AI-powered research assistant for analyzing food desert related documents.

Installation & Setup

1. Install Dependencies with Poetry

Install Poetry if you haven't already:

pip install poetry

Then, install the dependencies:

poetry install

2. Environment Setup

Create a .env file in the root directory and add your API keys:

OPENAI_API_KEY=your_api_key_here

3. Data Ingestion

Before running the application, you need to process the documents:

python ingestion.py

This will create a .chroma directory containing the document embeddings. Note that this directory might be large depending on your document collection.

Running the Application

Launch the Streamlit application:

poetry run streamlit run main.py

The web interface will automatically open in your default browser. You can:

• Enter your research question about food deserts
• Get AI-generated answers based on the document collection
• View relevant source documents with page references

Features

• Document relevance grading
• AI-powered answer generation
• Clean web interface using Streamlit
• Source document tracking with page references

Project Structure

.
├── main.py             # Streamlit application entry point
├── graph/              
│   ├── chains/         # LangChain components
│   └── nodes/          # Processing nodes
├── poetry.lock         # Poetry dependency lock file
└── pyproject.toml      # Project configuration