Every week includes hands-on Python work. Readings and materials are posted to Canvas. This page shows the topic, module, and deliverable due before each class session.

Start with Week 0 - account and software setup to complete before our first meeting.

Fall 2026 calendar

This course starts on a Wednesday and ends on a Monday, so the 11 calendar weeks of the quarter don't map one-to-one onto the 10 content units. Next time this course starts on a Monday, each content unit's two sessions will fall in the same calendar week.
Content unit Session 1 Session 2
Week 1 Class 1 - term Week 1 (only meeting) Class 2 - term Week 2, first meeting
Week 2 Class 3 - term Week 2, second meeting Class 4 - term Week 3, first meeting
Week 3 Class 5 - term Week 3, second meeting Class 6 - term Week 4, first meeting
Week 4 Class 7 - term Week 4, second meeting Class 8 - term Week 5, first meeting
Week 5 Class 9 - term Week 5, second meeting Class 10 - term Week 6, first meeting
Week 6 Class 11 - term Week 6, second meeting Class 12 - term Week 7, first meeting
Week 7 Class 13 - term Week 7, second meeting Class 14 - term Week 8, first meeting
Week 8 Class 15 - term Week 8, second meeting Class 16 - term Week 9, first meeting
Week 9 Class 17 - term Week 9, second meeting Class 18 - term Week 10, first meeting
Week 10 (presentations) Class 19 - term Week 10, second meeting Class 20 - term Week 11 (only meeting)

Module overview

Understand weeks 1-2  |  Use weeks 3-6  |  Evaluate weeks 7-8  |  Govern + Integrate weeks 9-10


Wk 0
Due before Week 1 begins: accounts, software install, setup check output, and pre-course survey | start here →
Wk 1
Deliverable: use case reflection with quantitative risk score | Python: first API calls, batch processing
Wk 2
Deliverable: tool setup and benchmarking report | Python: tokenization, model benchmarking | setup guide →
Wk 3
Deliverable: directed report with methods section and statistical test | Python: pandas cleaning at scale
Wk 4
Deliverable: debugging and review log with quantitative impact | Python: bug impact analysis, automated test suite
Wk 5
Deliverable: analytic design critique + implementation | Python: k-means with holdout validation or linear mixed models
Wk 6
Deliverable: text classification pipeline | Python: sklearn baseline vs. LLM classifier, inter-annotator κ, co-occurrence networks
Wk 7
Deliverable: prompt and output evaluation memo | Python: rubric scoring, multi-rater Cohen's κ, minimum detectable difference
Wk 8
Deliverable: AI incident brief | Python: subgroup disparity metrics, quantitative harm estimation
Wk 9
Deliverable: governance memo | Frameworks: NIST AI RMF, likelihood × impact risk matrix
Wk 10
Deliverable: final workflow package + 10-minute presentation | Includes peer evaluation of presentations