DataCareerHub
Explore learning
← Portfolio galleryFree companion · 45 minute suggested activity

Portfolio review: can another person reproduce it?

Turn an attractive project page into an inspectable analytical handoff with clear limits and practical next steps.

Learn → Try → ExplainReviewed 2026-11-01
01

The review scenario

Your project has a screenshot and a final number, but a reviewer cannot find the source, the query or the metric definition. Use the Harbor case or your own lawful dataset. The goal is a reproducible learning artifact, not a production deployment claim.

02

Build the handoff

Prepare a README with the question, fictional/source notice, schema, metric contract, setup steps, exact execution command, expected output, limitations and troubleshooting. Include the query and fixture separately. Record the environment you actually tested; do not list platforms you have not tried. Keep results separate from raw inputs.

03

Worked acceptance check

A reviewer starts with a fresh scratch SQLite database, loads the six-order fixture, runs the repaired query and obtains five orders, three customers, gross 410, refunds 35, net 375 and net per order 75.0. Then the equal-price order raises net to 435 without disappearing into DISTINCT. Describe these checks explicitly rather than saying “fully tested.”

04

Review accessibility and usefulness

Give charts a descriptive title, unit, date window and text explanation; do not rely only on color. Provide readable tables for key values. Explain the decision the project informs and what it cannot establish. Remove broken links and ensure a screenshot does not expose real identities, tokens or unpublished employer work.

05

Peer review protocol

One person follows the README without verbal help; the other records ambiguous steps and mismatches. Classify findings as correctness, reproducibility, communication or optional improvement. Fix the highest-impact issue, rerun the failing step and record the remaining limitation. Review the work, not the person.

06

Deliverable and next step

Create a short before/after review note: issue, evidence, change and retest result. Score setup, outputs, validation, limitations and presentation 0–2 each. If you later publish the project, label it as a learning project and state any assistance used. No exercise here grants a qualification or guarantees employment.

Make the reasoning yours.

Try the activity before reading its reference. Keep your sources, commands and assumptions with your work. Companion activities are unsaved self-practice and do not award a certificate or change lesson progress.

Resume guidance and examples are for learning and preparation only. They do not guarantee interviews, a job offer or employment.

These original companion activities are free. Advanced offline packs are separate supporter extras; companion activities do not change saved course progress.

Learner dashboard ↗ · Ask a question ↗