AI Development Workflow – Assignment Project
This repository contains the full submission for the AI Development Workflow Assignment under the AI for Software Engineering course. The project follows the complete end-to-end lifecycle of an AI system — from problem definition to deployment and monitoring — using a healthcare readmission prediction case study.
The work is organized into clearly structured parts to make navigation easy for reviewers, instructors, and collaborators.
📁 Repository Structure
AI_Development_Workflow_Assignment/
├── Part1_Short_Answer/
│ └── Part1_Short_Answer.md
├── Part2_Case_Study/
│ ├── Part2_Case_Study.md
│ └── Code/
│ └── part2_model_pipeline.py
├── Part3_Critical_Thinking/
│ └── Part3_Critical_Thinking.md
├── Part4_Reflection_Diagram/
│ ├── Part4_Reflection_Diagram.md
│ └── workflow_diagram.png
├── PDF_Report/
│ └── AI_Workflow_Assignment.pdf
└── README.md
📌 Overview of the Assignment Parts
🔹 Part 1: Short Answer Questions
Covers foundational concepts including:
- Problem definition
- Data sources & preprocessing
- Model development steps
- Evaluation & deployment considerations
🔹 Part 2: Case Study Application
A full hospital readmission prediction workflow including:
- Problem scope
- Data strategy & feature engineering
- Model choice and development
- Confusion matrix + precision/recall (hypothetical example)
- Deployment plan + compliance with healthcare regulations
- Overfitting mitigation strategies
Includes the main code file:
part2_model_pipeline.py
🔹 Part 3: Critical Thinking
Examines:
- Ethical concerns and data bias
- How bias impacts healthcare outcomes
- Trade-offs between interpretability, accuracy, and compute constraints
🔹 Part 4: Reflection & Workflow Diagram
Contains:
- Personal reflection on the workflow