I created an end-to-end ML project template that uses Prefect as the orchestrator 🚀 https://github.com/MatsMoll/ml-kickstarter This template is optimised for local development, in order to provide the fastest feedback loop as possible. Meaning it refreshes all pipelines, and models on save or promotion. Furthermore, it provides out of the box LLM / embedding servers, training pipelines, experiment tracking, pipeline orchestration, data catalogs, data lineage and much more!