For the AI Dev Tools Zoomcamp, a certificate of completion will not be issued unless, in addition to completing and submitting the final project, the learner also completes three peer reviews.
I am currently on peer review number 1. Ragulrao07 has developed a household chore manager.
I scored the project against the criteria found at AI Dev Tools Zoomcamp - Final Project
5. Application: Nothing ever runs python manage.py migrate. It is not in the Dockerfile, the compose file, or any entrypoint. A fresh docker compose up therefore starts against an empty PostgreSQL database with no tables, and the API should fail as soon as it tries to read data.
8. Deployment: The “deploy” jobs in the CI workflow are just echo statements: the comment even says they simulate pulling and running the container. Pushing an image to GHCR is a build artifact, not a deployment. The README also does not describe how to deploy to a real host.
9. CI/CD: Both deploy stages only print text, so nothing is actually deployed. What is real is that tests run automatically on every push to main (strictly speaking not on pull requests, which is a minor point).
11. Alerting: The Prometheus container mounts only prometheus.yml. The rules file is never copied into the container, so Prometheus cannot load the rule and the alert can never fire. The README also mentions Alertmanager, but it is not in the compose stack.
12. On-Call Agent: (1) As shipped, the alert cannot fire (see above), so the agent can never actually wake up on an alert. (2) There is no demo, log, or recording of it working. (3) Commit 2df80a8 uses a fix(on-call) prefix, but its message differs from the one the script tells the agent to use.
Add a migrate step to the container start, mount alert_rules.yml into Prometheus, attach an exporter to the TracerProvider, and replace the echo deploy jobs with a real deployment target.