Your course logbook

You fill this in live, with me, during each session. There is no separate homework worksheet: the work happens together in the room, and this is where you record it.

Make your own copy first

👉 Make a copy of the logbook and keep it open every session.

We reuse this later: in Workshop 6 you compare against your Workshop 1 baseline, and in Workshop 8 you build your final proposal straight from these notes. So keep the one doc for all eight weeks. Anything past the live work is optional polish, never required to keep up.

The headings below mirror the doc. Record results as we go.

---

W1 · Baseline

Pipeline timings from npm run ci (read them off baseline.log):

Stage	Duration	Pass / Fail
install
lint
typecheck
test:api
test:web
build:api
build:web
deploy
Total

• Biggest single stage (the hotspot): ______
• Two or three things that looked slow, redundant, or confusing: ______

W2 · Profiling and root cause

Your top three bottlenecks, ranked, each traced to a root cause (the end of your 5 Whys):

1. Bottleneck: ______ → root cause: ______
2. Bottleneck: ______ → root cause: ______
3. Bottleneck: ______ → root cause: ______

• Which one would you fix first, and why (impact vs effort): ______

W3 · CI/CD pipeline

Pipeline duration as we apply each optimization:

After step	Pipeline time	Note
Baseline
Caching
No always-on coverage
Parallel workers
Matrix + shard
(W5) mock the clients

• The honest finding: which steps moved the needle, and which barely mattered: ______

W4 · Developer experience

• Onboarding pain points we found on a fresh clone: ______
• Fixes we applied (setup script, README, seed step, Makefile, env keys): ______
• Time to a working app, before vs after: ______

W5 · AI-assisted development

For each task, what the AI produced and how it held up:

• Refactor of reports.ts: ______
• Generated tests for auth.service.ts: ______
• Mocking the SDK clients (and the test-speed change): ______

Cognitive-offloading check (the point of the session):

• I wrote my own comprehension-preserving system prompt
• Where I nearly accepted output I did not understand: ______
• I proved I understood the result by (explaining it back / modifying it without AI): ______

W6 · Experimentation and metrics

• My hypothesis (what change, expected effect, by when, why): ______
• Metrics I picked (DORA: deployment frequency, lead time, change failure rate, MTTR): ______
• Baseline vs improved (pull the W1 numbers from above):

Metric	Baseline	Improved

• Rollout plan in one line: ______

W7 · Standards and AI debt

• CI/CD standard: ______
• AI usage policy (one rule that matters): ______
• Coding standard: ______
• The one enforcement I actually implemented (hook / lint rule / PR template): ______

W8 · Final proposal

Outline, built from everything above:

• Problem (the slow, flaky pipeline, in numbers): ______
• Diagnosis (the bottlenecks and their causes): ______
• Changes I made (CI, DevEx, AI): ______
• Measured impact (before vs after): ______
• AI evaluation (where it helped, where it did not): ______
• Recommendation to leadership, in one sentence: ______