Free Model Monitoring Learning Path
The drift watch

GOAL
Create a local monitoring report for missing data, feature distribution and labeled prediction error, with explicit alert thresholds.

START
Read workbook.html, inspect case-data.csv, complete the five stage outputs, then the capstone. Use evidence-log.csv to record only actions actually performed.

CASE
Two synthetic batches each contain 100 scored requests. The reference batch has 5 missing ages, 20 premium-category records and 10 errors among 80 labeled records. The current batch has 12 missing ages, 35 premium records and 18 errors among 60 labeled records. Labels are delayed and incomplete.

TOOLS
Basic Python and percentages. Use CSV files and the standard library; no monitoring server or cloud account is required.

LOCAL COMMANDS / OPENING
python monitor.py

ASSESSMENT
Score at least 80/100, supply every required artifact, and satisfy the path-specific critical condition. Rework missing evidence before marking complete.
Critical condition: Unlabeled records are not treated as correct predictions and alerts do not claim a proven root cause.

SOURCE COURSES
MLOps Zoomcamp: Free Course — https://wikifree.com/courses/free-datatalks-club-mlops-zoomcamp/

RIGHTS AND SCOPE
Original WikiFree teaching case and exercises. All case data are synthetic. Provider lessons are linked, not redistributed. No certificate, real project delivery or unperformed runtime test is claimed.
