Most Resume Builders Start With a Template. Mine Starts With an Experience Library.

Every resume builder I tried had the same mental model:

Open a template. Fill it in. Export a PDF. Do it all over again for the next job.

That model quietly punishes you for applying to more than one role. The moment you want a "backend engineer" version and a "tech lead" version of your resume, you either duplicate the whole document or re-type your work history from scratch. Your best bullet points live in three different PDFs, none of which is the source of truth.

So when I built my job-application suite, I inverted the model:

The resume is not the primary object. Your experience is.

Here is why that distinction matters, and how the rest of the tool is designed around it.

The problem: resumes are documents, but careers are databases

A normal resume tool treats your career as one flat document. But your career is really a set of reusable facts:

• Jobs you've had (with highlights and the skills each one produced)
• Projects you've shipped
• Education, certificates
• Skills, both extracted and manually added
• Personal/contact profiles for different markets

If those facts are duplicated inside every resume, they drift. Update your current job in one PDF and the other five are now stale.

The fix is a normalized experience library: write each fact once, then compose resumes from it.

How the experience library works

Every piece of your background is stored once, as its own record. Work experience even carries its own skills, extracted per role:

model Experience {
  id          String
  title       String
  company     String?
  startDate   DateTime
  endDate     DateTime?
  description String
  highlights  String[]
  skills      String[]
  // ...
}

And each experience produces its own skills, so a resume can pull "the skills that came from this specific job" instead of a global, undifferentiated skill soup:

model ExperienceSkill {
  id           String
  name         String
  experienceId String   // skills are tied to a role
}

A resume never owns an experience. It references one, through a join table:

model ResumeExperience {
  resumeId     String
  experienceId String   // same experience, many resumes
  sortOrder    Int
}

That single design choice is the heart of the whole tool: one experience, many resumes.

Composing a resume = selecting, not retyping

Because everything is in the library, building a new resume becomes a selection exercise instead of a writing exercise.

In manual mode, you just tick the boxes:

• Which work experiences to include
• Which projects
• Which education and certificates
• Which skills (auto-extracted ones show their source role)
• Which personal fields to display (toggle per application)
• Drag to reorder

// A resume is assembled from selected library items,
// not typed from scratch.
const resumeData = {
  title: 'Frontend Engineer — Berlin',
  experience: selectedExperiences.map(toExperienceBlock),
  projectExperience: selectedProjects.map(toProjectBlock),
  skills: selectedSkillNames,
  education: selectedEducations,
  // ...
}

So a "frontend-heavy" version and a "leadership-heavy" version of your resume come from the same source of truth. Apply to five companies, tweak which highlights you surface for each, and never retype your career again.

There's also a profile concept on top: different contact/summary snapshots for different markets, so your China-market resume and your overseas resume don't have to fight over one header.

The second differentiator: templates that actually fit different industries

Most tools give you a handful of templates that are all basically the same layout with a different accent color. That is fine until you're applying to a Big 4 firm with a vaporwave resume, or to a game studio with a beige corporate one.

I went the opposite direction: templates with genuinely different design philosophies, mapped to real industries.

A few examples from the set:

• Classic / Minimal — clean and conservative for traditional industries
• Elegant / Finance / Consulting / HR / Sales — structured, professional layouts for finance, consulting, and client-facing roles
• Tech / Midnight Hacker — code-forward, terminal-style for developers
• Academic — the format reviewers expect for PhD and faculty applications
• Executive / Startup — leadership presence vs. high-energy early-stage energy
• Creative / Y2K Pop / Vaporwave / Glassmorphism / Kawaii / Sticker Cards / Swiss Grid / Pixel Quest / Front Page / Timeline — for design, marketing, and creative roles where standing out is the point
• Sidebar / International / One-Page — structure-oriented formats for personal-info-heavy, overseas, or condensed applications

{
  key: 'finance',
  name: 'Finance',
  description: 'Professional and stable, perfect for Big 4 and investment banking',
},
{
  key: 'midnight',
  name: 'Midnight Hacker',
  description: 'Dark terminal style with neon glow for developers',
},
{
  key: 'swiss',
  name: 'Swiss Grid',
  description: 'Bauhaus Swiss style with bold red-black contrast',
},

The point is not "more templates." It's that the right template is industry-dependent, and a designer, an investment banking analyst, and a PhD applicant should not be choosing from the same three navy-blue layouts.

Because the content comes from the same library, you can also take the same experience data and render it through a totally different template in one click — the library and the presentation are decoupled.

Why decoupling content from presentation matters

Once the library is the source of truth and templates are just views, a few good things fall out:

1. Tailoring is cheap. New resume per application? Select, reorder, export. No re-typing.
2. Updates propagate. Fix a typo in an experience and every resume built from it is correct.
3. Template choice becomes a strategy, not a lock-in. Same facts, different industry style, instantly.
4. Consistency stays high. Skills are tied to roles, so you don't accidentally claim a skill on a resume where it doesn't belong.

AI is a helper, not the model

There's AI generation too — you can paste a job description and let it draft, or you can go fully manual and compose from your library. But the important part is the architecture underneath: the AI writes into the same library-and-compose model, not into a throwaway document. You stay the owner of your experience data.

Try it

If you want to see the experience-library approach in action — reusable experiences, composable resumes, and a wide range of industry-specific templates:

Open the resume builder

You can build one resume, then fork it into several tailored versions from the same library, each with a template that fits the industry you're targeting.

The takeaway

Most resume builders are document editors with a template gallery bolted on.

This one treats your career as a structured, reusable dataset, and treats the resume as one projection of it — with templates that actually differ by industry instead of just by color.

If you're juggling multiple versions of your resume, the fix usually isn't "write better bullet points." It's "stop storing your career six times."