How are intermediate results handled between pipeline layers?

Each layer persists its output to a dedicated Workspace as intermediate assets, ensuring that nothing is recomputed unless explicitly requested and providing a clear audit trail.

What does the 'cooking metaphor' represent in this skill?

The metaphor organizes the system: the Kitchen is the Pipeline class, the Chef is the domain function, the Recipe is the YAML config, the Dish is the data asset, and the Academy contains the builder scripts.

Why is the haifn directory marked as 'DO NOT EDIT'?

The haifn directory contains auto-generated production functions. To change them, you must edit the builder scripts in code-dev/1-PIPELINE/ and run them to regenerate the production code.

What is the primary purpose of the haipipe system?

haipipe is a healthcare AI platform that uses a 6-layer sequential pipeline to transform raw clinical data into standardized datasets and eventually deployed ML models.

haipipe Data Pipeline Overview

Name: haipipe Data Pipeline Overview
Author: jluo41

byjluo41

0•

Data Science & ML

Provides a comprehensive architectural overview and navigation hub for the haipipe healthcare AI data pipeline system.

This skill serves as the central design reference for the haipipe system, a 6-layer sequential data pipeline designed for healthcare AI research and model incubation. It guides users through the platform's unique cooking-metaphor architecture—mapping components to Kitchens, Chefs, and Recipes—while explaining the relationship between core frameworks, ML models, and generated production functions. It is essential for understanding how to transform raw clinical data into deployable machine learning endpoints while maintaining strict schema consistency and human-AI collaboration standards.

Key Features

010 GitHub stars

02Builder pattern for automated generation of production-grade domain functions

036-layer sequential data pipeline architecture (Source to Endpoint)

04Config-driven execution using YAML and a dynamic @ reference system

05Standardized cooking metaphor for intuitive pipeline component management

06Unified asset management for persistent intermediate data storage

Use Cases

01Architecting end-to-end healthcare AI pipelines from raw clinical data to deployment

02Navigating complex academic research codebases for clinical ML model development

03Ensuring schema consistency and reproducibility across multi-stage data processing workflows

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jluo41/research-skills haipipe-data-0-overview

For use in Claude.ai and ChatGPT

Download Skill