Career and Individual Retrospective
Fall and winter, your major individual document looked forward: an RFC arguing for a decision. Spring’s looks back and then out: what did you build across this year, what did you learn about your own judgment, and how does that become career capital? This is the document where a year of reviewed pull requests, defenses, and decisions gets converted into things you can say in an interview and prove.
The team Project Retrospective is the shared raw material; this document is yours alone and goes deeper. Everything in it must be verifiable against the repo record, and your third Defense may probe any claim in it.
What You Must Produce
Section titled “What You Must Produce”A reflection document (2 to 3 pages, PDF) plus your updated resume (PDF), with peer review evidence included in the reflection.
1. Individual PR Portfolio
Section titled “1. Individual PR Portfolio”The 3 to 5 contributions from the whole year you are proudest of, each with a link or ID (students on NDA teams list IDs and include a sanitized excerpt) and a short caption: what problem it solved, why the approach was right, and what it demonstrates about your skills. These are the things you would walk an interviewer through. Pull requests are the expected currency, but an RFC that changed the team’s direction, a test strategy, or a deployment pipeline you own all qualify, as long as the repo record shows you drove it.
2. Judgment and Responsible Practice
Section titled “2. Judgment and Responsible Practice”A reflection on how your engineering judgment developed, grounded in specifics rather than growth language. Required content:
- A decision you made under uncertainty and how you would make it now.
- One concrete place an AI tool misled you this year and how you caught it: the wrong suggestion, plausible-but-broken code, or a confident false claim, plus the verification habit that caught it (or the failure that taught you the habit). After a year of AI-assisted engineering, “it never misled me” is not a credible answer; it means the verification was not happening.
- Where the responsibility boundaries you worked under (confidentiality, licensing, data handling, honest reporting to your partner) actually constrained a choice you made.
3. Career Artifacts
Section titled “3. Career Artifacts”- Intent versus outcome: your three fall intent lines, quoted verbatim from your Resume and Intent submission, each marked earned, changed, or dropped. An earned line cites its evidence (a pull request, a release, a metric); a changed or dropped line gets one sentence on what happened instead.
- Updated resume with the capstone as a real entry: title and dates, a one-sentence summary (with the project outcome if it strengthens the entry), 3 to 5 impact-oriented bullets describing your contributions with honest quantification, a technology line, and a link to the Landing Page or repository. One page unless substantial prior experience justifies two; ATS-friendly formatting.
- Peer review evidence, both directions: you give one structured resume review to a peer and receive one. Include the review you gave (specific, actionable, professional) and a short note on what you changed in response to the one you received. Swaps are arranged in week 8; reviews and revisions land in this submission.
Your cohort meeting with an instructor, by week 8, is required before this assignment is graded. It is the same cohort as in fall; the slot is published in Canvas in week 6, and the meeting walks the intent lines against the record. Missing it without rescheduling in advance holds the grade at zero until it happens.
Rubric (100 points)
Section titled “Rubric (100 points)”| Criterion | Points | Outcome |
|---|---|---|
| PR portfolio: 3 to 5 contributions, verifiable against the repo record, each with a caption that demonstrates a specific skill an interviewer would care about | 30 | SO2 |
| Judgment and responsible practice: specific, honest, includes a real AI-misled episode with the verification that caught it, and a real responsibility constraint | 25 | SO4 |
| Resume: capstone entry honest and impact-oriented, quantified where the numbers are real, professionally formatted, linked | 10 | L10 |
| Intent versus outcome: the three fall lines quoted, each marked earned, changed, or dropped, with evidence for the earned ones and one sentence for the others | 10 | L10 |
| Peer review: the review you gave is specific and actionable; the one you received produced a visible revision | 15 | L10 |
| Writing: precise, professional, within 2 to 3 pages, claims stated in verifiable language | 10 | SO3 |
AI use: AI may draft prose and polish resume bullets; disclose tools and verification as usual. Fabricated contributions, inflated metrics, or an invented AI-misled anecdote are academic dishonesty on the one assignment explicitly about honest self-assessment. Everything here is checkable, and it will be checked.
Activities That Prepare This
Section titled “Activities That Prepare This”| Activity | Criterion it serves |
|---|---|
| Portfolio Development | PR portfolio |
| Public Code Repository | PR portfolio |
| Personal Branding | PR portfolio captions |
| Mock Interviews | PR portfolio |
| Resume Building | Resume, and the peer-review exchange |
| Learning Journal | Judgment and responsible practice |
| Learn from Jean-Luc Doumont | Writing |
Two of these do more than their line suggests. Resume Building already includes a peer-review round, which is the same exchange this assignment’s peer-review criterion grades: do it there first and the graded version is a revision rather than a first draft. Personal Branding asks you to write up one real problem you solved, which is the same move each PR portfolio caption has to make in three sentences.
Mock Interviews is the cheap way to find out whether your captions survive contact with someone who was not there. The PR portfolio criterion asks for contributions “you would walk an interviewer through”; a mock interview is that walk, done early enough to fix.
Browse career and reflective when these run out.