Phase 4: Autonomous Workflows

Design full AI systems that run with minimal human supervision

Modules in This Phase

Error Handling & Logging

Professional AI systems don't crash silently — they handle errors gracefully and log what happened. Learn to build resilient, debuggable workflows.

Duration

Week 37–39

Tools

PythonSQLiteGitHub

Projects

2 hands-on projects

Milestone: Your AI workflow logs every decision and handles errors without crashing.

Evaluation Metrics

How do you know if your AI system is actually good? Learn to define metrics, measure output quality, and use data to improve your systems over time.

Duration

Week 40–42

Tools

PythonGoogle SheetsOpenAI API

Projects

2 hands-on projects

Milestone: You can measure and compare the quality of two different AI approaches.

Version Control with GitHub

Real engineers use GitHub to track changes, collaborate, and build portfolios. Learn Git basics and publish your AI projects for the world to see.

Duration

Week 43–45

Tools

GitHubGitVS Code

Projects

2 hands-on projects

Milestone: You have at least 3 AI projects published on GitHub with clear documentation.

Capstone: Autonomous Academic Agent

Put everything together! Build your final capstone project: a fully autonomous AI agent that accepts a learning goal, creates a study plan, tracks progress, and generates weekly reports.

Duration

Week 46–48

Tools

PythonOpenAI APISQLiteGitHub

Projects

1 hands-on project

Milestone: You present a working autonomous AI agent with planning, execution, self-review, and logging.