Phase 10: Career, Portfolio & Interview Engineering · 45 min · Claude Code · Python

Resume, LinkedIn & Job Search Strategy

A resume stuffed with every AI buzzword gets filtered by the same semantic ATS matching it was trying to game. A resume describing your actual spec, harness, and quality gate — in the job posting's own specific language — passes the filter and is also just true.

Hiring signal: Tailoring a resume by honestly mirroring a posting's specific technical language with real portfolio evidence is the same specificity discipline this course has taught for specs and postings, now applied to your own job search materials.

What you will learn

Explain how 2026 ATS systems use semantic and contextual matching, making unearned keyword stuffing actively counterproductive
Mirror a target posting's specific technical language honestly, backed by real portfolio evidence rather than unsupported buzzwords
Target resume and LinkedIn positioning by the company archetype (from c12-10-1) a specific posting represents
Present your course-long product's full three-orbit arc as the centerpiece of a job search narrative

Introduction

Resume, LinkedIn & Job Search Strategy

A resume lists "AI agents, LLMs, RAG, prompt engineering, LangChain, vector databases, fine-tuning" in a dense skills block at the top, repeated in slightly different words throughout. It gets filtered out before a human ever opens it. The posting it was submitted against used specific language — "spec-driven development," "harness engineering," "drift-gate CI enforcement" — that never appears anywhere on the resume, even though the candidate has built exactly that system across this course. A second candidate, applying to the same posting, writes one bullet: "Designed a spec-driven refund feature with a harness enforcing tool-access limits and a drift-gate CI check blocking unreviewed spec/code divergence." It mirrors the posting's actual language, and it's also completely true, backed by a real repository. The first resume was optimized for keywords it thought the job wanted. The second was optimized for honestly describing what was actually built, in the vocabulary the posting itself used — and it passed the same filter the first one didn't.

How 2026 ATS filtering actually works

Modern applicant tracking systems (Workday, Greenhouse, iCIMS) have moved well past simple keyword counting into semantic and contextual matching — a skills-graph model that recognizes when a keyword is stuffed in without supporting context versus genuinely woven through real experience descriptions. Keyword stuffing isn't just ineffective anymore, it's actively counterproductive against these systems, while a resume whose bullets naturally use a posting's specific terminology because that's genuinely what was built matches well precisely because the match is real, not manufactured. This is the same distinction c12-10-1 drew between buzzword-heavy and signal-heavy job postings, just applied in the opposite direction: your resume's specific, checkable language is what both a semantic ATS and a human reviewer are actually looking for.

Mirror the posting's language honestly, don't invent experience to match it
Rewriting a bullet to say "designed a drift-gate CI enforcement pipeline" is honest mirroring if you actually built one — it's the same artifact, described in the posting's vocabulary instead of your own. Claiming that language for work you didn't do is fabrication, and it fails the same way a fabricated portfolio claim does: an interview will eventually ask you to go deeper than the resume line, and there won't be anything real underneath.

Unlock the full lesson

You've read the first 2 sections. The rest of this lesson covers Targeting by company archetype, not one generic resume, The demo-day presentation: your product's full arc, LinkedIn as a standing version of the same evidence, Build It — plus a hands-on lab, quiz, and project artifact.

Create a free account to unlock Phase 0 and Phase 1 of every course — no credit card.