I Rebuilt My Resume, LinkedIn, and Website in a Weekend with Claude

This post was written with AI assistance.

I hadn’t touched my resume in years. My LinkedIn was stale. My personal website was a template I deployed once and forgot about. All of it was out of date, disconnected, and didn’t reflect where I actually am in my career.

So I rebuilt the whole thing in a weekend using Claude Code. Resume, LinkedIn, personal website, blog — all of it. Here’s how.

The problem

My professional presence was scattered. My resume said one thing, my LinkedIn said another, and my website said nothing useful at all. Every time I needed to update one, I’d open a Google Doc, tweak a few lines, and forget about the rest. Nothing was connected.

I knew I needed to fix it, but the thought of manually updating three different formats with consistent messaging felt like exactly the kind of task I’d keep putting off. So I didn’t do it the old way.

The setup

Before writing a single line of content, I set up a project folder with everything Claude would need to work with. The quality of the output depends entirely on the quality of the input — so I was deliberate about this.

resume-update/
  content/              # Markdown source files
    resume.md           # The single source of truth
    cv.md               # Full career history
    linkedin.md         # Copy-paste ready LinkedIn sections
  career-notes/              # Years of personal career notes and self-assessments
  jessepeplinski.github.io/   # Personal website (Astro)
  CLAUDE.md             # Project instructions for Claude Code

The key file is CLAUDE.md. This is where I told Claude what I was working on, where the source materials lived, my official job title, and that the content should match my writing voice. Think of it as a project brief that lives in the repo — Claude reads it automatically every time you start a session.

The content pipeline

The whole system runs off one file: resume.md.

resume.md → LinkedIn profile
resume.md → Website content
resume.md → Print-friendly PDF (via Astro page + Cmd+P)
cv.md     → Full historical record

When I update resume.md, everything downstream can be regenerated from the same source. No copy-pasting between Google Docs and LinkedIn. No wondering which version is current. One source of truth.

The career notes trick

This was the move that made the biggest difference. I fed Claude years of personal career notes and self-assessments. These had specific projects, metrics, and reflections that I’d completely forgotten about — things that never made it onto my old resume because I wrote that resume years ago and never went back.

Claude surfaced patterns I couldn’t see myself. It connected dots across years of notes and helped me articulate impact I knew was there but hadn’t put into words. Things like leading a major platform migration, driving a page that generates significant business value, and shipping a customer portal — all buried in notes I hadn’t read in years.

The raw material was always there. I just needed a way to process it.

Building the website

The website is an Astro site with Tailwind CSS, deployed on Netlify. Claude built the entire thing — every component, every page, the build config, the deployment pipeline. I described what I wanted and we iterated.

What I directed:

Dark mode with an orange accent (my brand colors)
Clean, minimal layout — no template bloat
A print-optimized resume page so I could Cmd+P a PDF without touching Google Docs
A blog section for writing like this

What Claude built:

Reusable Astro components (Header, Footer, layout system)
Responsive design with Tailwind
SEO setup (structured data, Open Graph tags, sitemap, RSS feed, robots.txt)
The entire CSS and markup for a print-friendly resume that matches the site’s brand

I didn’t write the code by hand. I made the decisions, described what I wanted, reviewed every output, and course-corrected when something didn’t feel right.

The LinkedIn update

This one was simple but effective. I had Claude generate a linkedin.md file — copy-paste ready sections for every part of the LinkedIn profile. Headline, about section, experience bullets, all of it.

Because it was generated from the same resume.md source, everything was consistent. Same metrics, same framing, same voice. I just opened LinkedIn, pasted the sections in, and I was done. No agonizing over whether my LinkedIn headline matches my resume summary.

What I actually did vs. what Claude did

Me:

All the raw content, ideas, and career history
Design direction and brand decisions
Voice and tone — Claude adapted to mine, not the other way around
Final review and approval of everything
The decision of what to include and what to cut

Claude:

All code implementation (HTML, CSS, Astro, Tailwind, config)
Content refinement and organization
Surfacing patterns from years of career notes
Building the site infrastructure and deployment pipeline

What I’d tell someone starting from scratch

Gather your source material first. Career notes, self-assessments, old resumes, project notes, emails where someone said you did a good job — dump it all in a folder. The more context Claude has, the better the output.
Write a CLAUDE.md file. Tell it who you are, what you’re working on, and how you want the output to sound. This is the highest-leverage thing you can do.
Pick one source of truth. For me it’s resume.md. Everything else is a downstream artifact. This is what keeps things consistent when you inevitably update something six months from now.
Don’t try to write the final version first. Get the raw ideas down in your own words. Let Claude help you organize and refine. The thinking has to be yours — the polish is where AI shines.
Ship it. You could spend another week tweaking font sizes. Don’t. Get it live, iterate later.

The result

In a weekend, I went from a stale Google Doc resume and a forgotten website to a fully branded online presence — website, blog, resume, LinkedIn — all connected, all consistent, all deployed. The kind of thing that would have taken me weeks to do manually, if I ever got around to it at all.

That’s the thing about these tools. They don’t do the thinking for you. They make it possible to actually execute on the ideas you’ve been sitting on.