Insurance Policy Chatbot

Install local virtual environment for 'injection' service

Create the virtual environment for the 'injection' service by running:

python3 -m venv injection-env

Activate the virtual environment by running:

source injection-env/bin/activate

Install the required packages by running:

pip install -r injection/requirements.txt

Install the spacy model for spanish language by running:

python -m spacy download es_core_news_sm

Build and run the Qdrant service

When you're ready, start the vector database by running:

docker compose up qdrant --build -d

Create the vector database collections

First, you have to create a .env file inside the injection directory. Follow the raw_env.txt guide. After creating the injection/.env file you can load the document/vectors collections, run the following files from the injection directory:

1. cd injection
2. python main.py

Build and run the Chat service

First, you have to create a .env file inside the backend_ui directory. Follow the raw_env.txt guide. After creating the backend_ui/.env file you can deploy the service container by running:

docker compose up backend_ui --build -d

Run the app

Your application will be available at http://localhost:8000.