Your resume, audited like code.

---

The problem

Most resumes never get read by a human. They get filtered by an ATS, skimmed for eight seconds, or quietly ranked below a hundred others — and nobody tells you why. ResumeLM is the diagnostic step that happens before you hit "Apply."

Upload a resume. Get back a score, a list of concrete issues, and a feedback report you can actually act on — not a vague "looks good!" from a friend who didn't really read it.

How it works

  upload          parse              score              fix
┌─────────┐    ┌──────────┐    ┌─────────────────┐    ┌──────────┐
│  .pdf   │ →  │ extract  │ →  │ ATS + Job Fit    │ →  │ targeted │
│  .docx  │    │  content │    │ scoring engine   │    │ feedback │
└─────────┘    └──────────┘    └─────────────────┘    └──────────┘

1. Upload your resume (PDF or Word)
2. ResumeLM parses structure, content, and formatting
3. It's scored against ATS-readability rules and, optionally, a specific job description
4. You get an itemized breakdown: what's hurting your score, what's missing, and what to fix first

What you actually get back

Output	What it tells you
ATS Score	A single number reflecting how cleanly an applicant tracking system can parse your resume
Issue list	Specific, line-level problems — not generic advice
Job Fit Score	How well your resume matches a specific job description, when you provide one
Missing keywords	Terms the job description expects that your resume doesn't have
Matching strengths	What's already working, so you know what to keep

No fluff scoring, no "9/10 great job!" — if something's broken, ResumeLM says so.

🔜 Coming soon: Cover Letter Generation

The next piece of the pipeline. Once your resume is scoring well, ResumeLM will generate a tailored cover letter that:

• Pulls directly from your scored resume — no rewriting your story from scratch
• Aligns tone and keywords to the specific job description you're targeting
• Comes back editable, not a black-box wall of text you have to untangle

This isn't on by default yet — it's actively in development and will ship as a natural extension of the existing scoring engine, not a bolted-on feature.

Stack

• Framework — Next.js (App Router, Server Actions)
• Language — TypeScript
• Auth — Better Auth, email/password with verification
• Database — PostgreSQL via Supabase, accessed through Prisma 7 with driver adapters
• API — Groq API
• State management — Zustand
• Hosting — Vercel

Getting started

git clone https://github.com/your-username/resume-lm.git
cd resume-lm
npm install

Create a .env with your database and mail credentials:

DATABASE_URL="postgresql://<user>:<password>@<pooler-host>:6543/postgres?pgbouncer=true&connection_limit=1"
NEXT_PUBLIC_APP_URL="http://localhost:3000"
EMAIL_FROM="noreply@yourdomain.com"

Generate the Prisma client and run migrations:

npx prisma generate
npx prisma migrate dev

Start the dev server:

npm run dev

Roadmap

• Resume upload + parsing
• ATS scoring engine
• Job-description-aware fit scoring
• Email/password auth with verification
• Cover letter generation
• Resume version history
• Export annotated PDF report

---

_{Built for people tired of guessing why their resume isn't working.}