Skip to content

35 - Capstone Blueprints (Step-by-Step Endgame)

Home: README

Path placeholder: <repo-root> means the folder containing this repository's README.md.

These are execution blueprints for final mastery-level deliverables.

Capstone 1 - Automation platform

Goal: - Build resilient file/Excel automation with validation, rejects, logs, and tests.

Execution steps: 1. Define input/output contract. 2. Build parsing and validation layer. 3. Build transformation layer. 4. Build output + rejects workflow. 5. Add tests and logging. 6. Add rerun safety checks. 7. Document runbook.

Command scaffold: 

cd <repo-root>/projects/level-5/15-level5-mini-capstone
python project.py --input data/sample_input.txt --output data/output_summary.json
pytest -q

Expected output: 

... output_summary.json written ...
2 passed

Capstone 2 - Data and SQL pipeline

Goal: - Build idempotent staging-to-reporting pipeline with lineage and drift handling.

Execution steps: 1. Design staging/reporting schema. 2. Implement ingestion + idempotency key logic. 3. Implement promote/merge logic. 4. Add quality checks and summaries. 5. Simulate schema drift and recover. 6. Document lineage and risk controls.

Command scaffold: 

cd <repo-root>/projects/level-6/15-level6-mini-capstone
python project.py --input data/sample_input.txt --output data/output_summary.json
pytest -q

Expected output: 

... output_summary.json written ...
2 passed

Capstone 3 - Full-stack operational system

Goal: - Build browser-consumable operational system with backend, data layer, and resilience.

Execution steps: 1. Define user stories and API contract. 2. Build backend endpoints and data access layer. 3. Build browser dashboard layer. 4. Add caching/fallback behavior. 5. Add tests + quality checks. 6. Add failure simulation and recovery steps.

Command scaffold: 

cd <repo-root>/projects/level-10/15-level10-grand-capstone
python project.py --input data/sample_input.txt --output data/output_summary.json
pytest -q
ruff check .
black --check .

Expected output: 

... output_summary.json written ...
2 passed
All checks passed!
would reformat 0 files

Final pass criteria

Pass when all are true: 1. All three capstones are reproducible from docs only. 2. Behavior is deterministic across reruns. 3. Failure modes are documented and validated. 4. Tests and quality checks are clean. 5. Architecture decisions and tradeoffs are clearly documented.

Fail when any are true: 1. Hidden setup assumptions remain. 2. Outputs differ without intentional changes. 3. Failure handling is not validated. 4. Tradeoff reasoning is unclear.

Primary Sources

Optional Resources

Next

Next: 36_ELITE_ENGINEERING_TRACK.md →