This repo contains a sample RAG system using hatchet to deal with the batch processing and execution.

This code will:

1. reads a website and parses the text content
2. reasons about what information is most relevant to the user request
3. generates a response for the user

This repo contains:

• Frontend react application
• FastAPI that will act as our edge service
• Hatchet worker

In order to execute this code you need to have:

• poetry
• npm
• Hatchet
• docker
• docker-compose
• Ollama

Before building and running the application make sure to start docker and create a API token,

Create a .venv file based on .venv-example including your generated token

In order to install required packages run the following command

poetry install

This will install the required packages

Start the API with

poetry run api

In a different Terminal start the hatchet worker with

poetry run hatchet

In another terminal go to the frontend folder and start with

npm start

You can now use http://localhost:3000 and test the query and http://localhost:8080 to view the execution status of the workflow.

This application is based on the tutorial https://docs.hatchet.run/home/tutorials/fastapi-react/project-setup with minor changes to use ollama.