Bookworm

Bookworm is a FastAPI based book management api with AI recommendation system (postgres as database). It's easy to deploy, manage and scale. As the project is solo maintained github copilot is used for reviewer for basic code reviews.

Prerequisites

Before you begin, you'll need to install one important software:

Install docker

Docker is required to run the application, visit docker for desktop and follow the steps.

Important

For the application to deploy .env.dev file should be configured properly. Check env/README.md before proceeding.

Running the application

The following commands will help developers to spin-up the bookworm app local dev server. You have two command options for running bookworm, using Docker or Make.

Option 1: Using Docker command

1. Clean-up (optional step, remove created volumes and containers if any)

docker compose down -v

1. Build and setup application

docker compose up -d --build

This command will do the following.

1. Build docker image and deploy api, db, ai and redis services.
2. Start up the database.
3. Run api using uvicorn in port 8000.

2. Pull model from ollama for ai service to use

docker exec -it bookworm-ai ollama pull llama3.2

Note

change the llm model name according to value provided in .env.* file default is set to llama3.2

3. Load sample data to db (optional step)

docker exec -i pg-db psql -U postgres books < utils/data/db_loader.sql

Option 2: Using Make command

1. Clean-up (optional step)

make destroy

2. Build and setup end to end application including local llm service.

make setup

Note

make setup is a collection of 3 targets. It builds the application first, the pull-model downloads the llm from the ollama hub and finally load_data load the db with some books data. Additionally a books_admin user also will be added out of the box.

Features to be added

• Use uv or poetry package for dependency management.
• Implement pgai for book recommendation.