How umaqgeek/apadilangit_workshop works
Project type and purpose
• This is an “AI, Data, and Aerospace” collection of focused data‐science/AI pipelines rather than a monolithic web app or microservices system.
• Each Python script under scripts/ implements one aerospace‐related workflow (flight‐delay prediction, satellite imagery analysis, rover navigation, generic data preprocessing).
• The core purpose is to demonstrate and run different AI/ML tasks on aerospace datasets, storing and loading trained models from models/.
File structure
• Top-level folders:
• datasets/ → raw inputs (CSV, JSON, images)
• scripts/ → individual Python workflows
• models/ → serialized/trained models (pkl, h5)
• Root config files: requirements.txt, pyproject.toml (dependency management).
• No single “orchestration” folder—each script is a self-contained pipeline.
Key insights from the README
• Installation via pip/Poetry.
• Clear one-to-one mapping: run script → process data → (train or infer) → output.
• No explicit service-oriented or front-end components—pure back-end data/ML pipelines.