Build evidence of your capability
Describe your contribution, AI assistance and limitations truthfully.
Reviewed: 2026-10-03
Illustrated teaching example • Fictional data
Narration: Microsoft Mark · locally synthesized

00:00:00.000 — Show what you can demonstrate.

Career evidence is stronger when a reviewer can inspect your work and understand your contribution. A certificate or polished claim alone does not show capability. We will repair an inflated description of a fictional project without hiding the useful work that actually happened.

Visual description:
Contribution · Artifact · Checks · Limits

00:00:19.333 — Start with the actual project record.

Riley wrote two tests using ten synthetic cases. A teammate built the interface. AI assisted with the README, which Riley reviewed. Two unsupported outputs were found. The project was not deployed to users, and savings were not measured. These facts define the scope of a truthful description.

Visual description:
Riley: 2 tests · 10 synthetic cases
Teammate: Built the interface
AI assistance: README draft, reviewed by Riley
Limitations: 2 unsupported outputs; no deployment

00:00:43.000 — Remove invented scope and outcomes.

The inflated description says Riley independently built and deployed a reliable system that saved forty percent. The evidence does not support sole authorship, deployment, reliability across real users, or measured savings. Correct those claims while preserving the testing and review contribution.

Visual description:
Independently built and deployed a reliable system; saved 40%.
Sole authorship: Teammate built UI (fail)
Deployment: No user deployment (fail)
Savings: Not measured (fail)

00:01:05.458 — Write a defensible project description.

A stronger description says Riley authored two tests over ten synthetic cases, documented two unsupported outputs, and reviewed an AI assisted README. It credits the teammate’s interface work and states that user deployment and savings are unmeasured. Specific contribution is more credible than inflated certainty.

Visual description:
Authored two tests covering ten synthetic cases; documented two unsupported outputs and reviewed an AI-assisted README. Interface built by a teammate. No user deployment or measured savings.

00:01:28.375 — Package evidence for a reviewer.

Include a sanitized artifact, test cases and results, your contribution, AI assistance, and known limitations. Explain how the reviewer can inspect or reproduce the work. Do not publish sensitive information, unsupported business claims, or another person’s work as your own.

Visual description:
Artifact: Sanitized, inspectable example
Checks: Cases, results and failures
Contribution: What you and others did
Limits: Unverified outcomes and next work

00:01:50.750 — Prepare your own evidence packet.

Use the lesson’s fictional project record or an authorized sanitized project. Write a concise contribution statement, attach the relevant artifact and checks, and name one limitation. Passing means a reviewer can distinguish demonstrated skill from unmeasured outcomes and understand where AI helped.

Visual description:
Could a reviewer verify your contribution?
Inspectable artifact.
Your role and AI assistance.
Checks, failures and limitations.
Your evidence: a truthful, reviewable project packet.