WAYNE RAINEY · THE CAREER CANTINA · 30 MIN
You scan the results, click a page, read it yourself. The human decides what's relevant.
Perplexity, ChatGPT Search, Google AI Overviews. The machine decides what's relevant — and cites its sources.
The question is no longer "does your page rank?"
It's "does the AI know you exist?"
So you can be found by systems and people
who don't know yet that they're looking for you.
Discoverability before the search. In the pool before the req is posted.
Ambient visibility without active effort. Networking at rest.
Credibility through demonstrated systems literacy. The repo is the proof.
Crawls the open web in real time. Constructs answers and cites sources explicitly. If your GitHub README is indexed, Perplexity can cite it when someone asks about you or your topic area.
OpenAI web search layer. Retrieves and synthesizes from indexed pages. Same architecture as Perplexity — open web only.
AI-generated summaries above organic results. Pulls from indexed content. LinkedIn profiles are partially indexed — GitHub has no crawl restrictions.
Built on Bing's index. Retrieval-then-synthesis. Same open web dependency.
Builds talent graphs from public web data including GitHub. Currently named in an FCRA lawsuit over how it processes that data. Your audience should know this tool by name.
Explicitly crawls GitHub profiles as a sourcing signal. An HR professional with a GitHub presence is findable in SeekOut. A LinkedIn-only professional is not.
Open web sourcing. Aggregates public professional data across platforms beyond LinkedIn.
Attributes-based talent search pulling from multiple public sources. More surface area = more attributes = more matches.
LinkedIn actively limits deep crawling. GitHub has no such restriction. The same professional — same experience, same skills — is findable in one and invisible in the other.
Get humans to click your link by ranking high for keywords. Heuristic. Rule-of-thumb. Pattern matching on surface features.
Get AI systems to cite your content when constructing an answer. Meaning-aware. Entity-based. The same two-stage architecture as AI hiring systems.
SEO is to AEO as keyword scan is to embedding match — sound familiar?
Everything you've built on LinkedIn serves LinkedIn's algorithm. It's optimized for their system, their crawlers, their data model. Not the open web.
A public, structured, machine-readable document that tells every answer engine, every crawler, every future system who you are — on your terms.
"LinkedIn owns your profile. The markdown file on GitHub is yours."
Answer engines don't search for strings of text. They search for entities — named things with attributes, relationships, and a body of evidence that confirms their identity. People. Organizations. Concepts. Frameworks.
named entity vs generic keyword
Inside the walled garden. Serves LinkedIn's algorithm. Essential but insufficient. The AI can't cite what it can't crawl.
Free. Browser-only. High-authority domain. Machine-readable markdown. Indexed fast. No domain required. Used by zero HR people. Yet.
Custom domain on Netlify, Squarespace, or similar. Aspirational but not required. GitHub gets you 80% of the value for 0% of the cost.
Your resume. Opt-in by design. Sits behind a gate. Only works when someone asks for it or you send it. Reactive. Transactional. One recipient at a time.
Your entity document. Ambient by default. Already out there. Discoverable before the request exists. Works while you sleep. Infinite recipients simultaneously.
GitHub is one of the most trusted domains on the internet. Your content inherits that authority the moment it's published.
Google visits GitHub frequently. Your entity document gets indexed within days — no Search Console required.
An HR professional with a GitHub repo signals systems thinking. It forces a reframe. That question is worth more than the answer.
An elevator pitch waits for the right room, the right person, the right moment.
A README delivers your pitch simultaneously to every crawler, every answer engine, every future system — permanently, while you sleep.
Describes mechanics. No entity signal. No semantic anchoring. Invisible to answer engines.
Establishes who, what, why. Names your frameworks. Links your properties. Reads like a Wikipedia entry that you wrote first.
Contact information
Summary and About section
Work experience and job titles
Education and degrees
Skills and endorsements
Certifications and licenses
Featured section and curated content
Volunteer work and community roles
Recommendations from colleagues
Publications and articles
Projects and portfolio work
Honors, awards, recognitions
The PDF LinkedIn hands you reflects its 2003 database schema. Everything added since to capture richer professional identity? Gone on export.
Featured exists because of LinkedIn activity. It is downstream proof of publishing, engaging, and building topical authority. The algorithm sees it and uses it. It never appears on a resume. It never survives the PDF export. The open web never knows it existed.
A resume might say Board Member. That tells the system almost nothing. The full story - what you built, who you served, what you presented, what changed because you were there - is where the semantic value lives. Most resumes compress or omit it. LinkedIn PDF export drops it completely.
Your GitHub entity document is the rescue operation for orphan data. Everything that does not survive the export belongs here - deliberately, explicitly, in your own words.
Yes, we just said the resume is pull media. We use it here because everyone already has one — it is the fastest path to structured professional history. The resume is the seed. What comes out the other side is something the resume could never be.
Your resume is already in your hands — Word doc or PDF both work. Everyone has one. No download process, no settings navigation required.
Claude extracts, enriches, normalizes your titles, surfaces implicit skills, and structures it for semantic discovery. You review and refine — it should sound like you.
Use a domain email if you have one — otherwise your cleanest Gmail. Username = your name, no numbers or underscores. Sign up at github.com — browser only, no installation.
New repo → public → add README → pencil icon → select all → paste → commit. You are live. Public URL is yours immediately.
Featured section or About section. Two high-authority indexed domains pointing at each other = entity confirmation signal for every answer engine that crawls both.
Three Layers of Source Material — Go as Deep as You Can
Resume and/or LinkedIn PDF
Available immediately. Zero wait. Does just enough to seed a solid entity document today. The LinkedIn PDF has orphan data gaps — your README fills them manually.
LinkedIn Archive: Profile + Recommendations
Settings → Data Privacy → Download my data. Select Profile and Recommendations only. Request archive — wait for delivery. Feed to Claude to enrich your README. High friction. High yield frosting.
Your Own Domain
Personal site on Netlify, Squarespace, or similar. Maximum entity coherence. Maximum surface area. Aspirational for most — but GitHub gets you 80% of the value for 0% of the cost right now.
Don't wait for Layer 2 or 3 to start. Build from Layer 1 today. Enrich later.
The choice isn't between doing it perfectly or doing it wrong. It's between a job seeker who publishes today and one who plans to do it properly — and never does.
Answer Engine Optimization. Optimize to be cited by AI, not just ranked by search. The game has shifted layers.
Same name, same email, same links across LinkedIn, your site, and GitHub. Consistency is how machines confirm identity.
More indexed nodes = more discovery paths. You're not building a website. You're building a knowledge graph about yourself.
Your homework: Five steps. One hour. One public entity document that works while you sleep.
Which situation describes you?
Same workflow as everyone else. Create a new public repo where the name matches your username exactly — yourname/your-name. Add your README. Done.
Even simpler. Open the repo, click Add a README, paste your entity document, commit. GitHub will prompt you if the file doesn't exist. Two clicks to start.
Your README is probably code documentation. Don't overwrite it. Instead, create a dedicated profile repository — a repo where the name exactly matches your GitHub username. GitHub auto-displays that README on your public profile. Your technical work stays untouched. Your entity document lives separately.