The AI-powered job search platform that works both ways.
Find the right jobs for your resume. Send the right resume for every job.
Most job search tools solve one problem: find jobs. Most resume tools solve the other: improve your resume. RACK solves both — simultaneously, intelligently, and with the same AI pipeline.
RACK is a two-sided matching platform:
- → You paste a job description. RACK ranks all your resumes by fit, explains exactly why each one scores the way it does, and tells you which to send.
- ← We fetch jobs for you. RACK automatically pulls fresh postings from top tech companies and ranks them against your profile and all the uploaded resumes in your profile — so the best opportunities surface, not just the newest ones.
No more guesswork. No more spray-and-pray. No more wondering if your resume is landing.
Traditional approach: RACK approach:
┌─────────────────────────┐ ┌──────────────────────────────────────────┐
│ Job Board → You apply │ │ RACK fetches jobs → AI ranks by fit │
│ with same resume every │ VS │ You paste JD → AI ranks your resumes │
│ time, hope for the best│ │ You always know the score, literally │
└─────────────────────────┘ └──────────────────────────────────────────┘
Paste any job description. RACK instantly scores all your uploaded resumes against it using a multi-phase AI pipeline — semantic search to shortlist, then deep LLM analysis across three dimensions:
- Skills Fit — how well your technical skills align with requirements
- Experience Fit — seniority and domain depth match
- Trajectory Fit — career arc alignment with the role's growth path
Each result includes key strengths, gaps, and a direct hire recommendation. You see the AI work in real time.
Set your profile once. RACK does the rest.
- Automatically scans top tech company job boards daily
- Ranks results by a blend of match quality and recency — freshness is a tiebreaker, not the gate
- Archive jobs permanently — they never resurface
- Curated signal, not firehose volume
- Unified view of Auto Matches and your saved Watchlist
- Archive, save, and organize postings
- Location-aware filtering
RACK uses a two-phase pipeline to balance speed and accuracy:
Phase 1 — Semantic Shortlisting Vector search narrows the candidate pool to only genuinely relevant matches, fast.
Phase 2 — LLM Deep Scoring GPT-4o-mini holistically evaluates each resume×job pair and returns structured scores, reasoning, and a recommendation tier.
Recommendation tiers:
| Score | Tier |
|---|---|
| ≥ 85 | 🟢 Strong Match |
| ≥ 65 | 🔵 Good Match |
| ≥ 50 | 🟡 Partial Match |
| < 50 | 🔴 Poor Match |
| Layer | Technology |
|---|---|
| Frontend | React 19 + Vite + Tailwind CSS |
| Backend | FastAPI (Python) |
| Vector Search | FAISS + sentence embeddings |
| LLM Scoring | GPT-4o-mini (OpenAI) |
| Job Data | Curated tech company job boards |
RACK is built for real-world use — including on your phone.
- Smooth animations and gradient UI
- Full iOS viewport fix (no browser chrome clipping)
- Cards expand cleanly, titles never truncate
- Bottom safe-area padding for notched devices
- Node.js 18+
- Python 3.10+
- OpenAI API key
cd rack/frontend
npm install
npm run devcd rack/backend
pip install -r requirements.txt
uvicorn main:app --reloadOPENAI_API_KEY=your_key_here- Outcome feedback loop — applied → interview → offer signals feed back into match ranking
- Expanded job sources
- Resume gap coaching based on match analysis
- Mobile polish across all pages
Recruiters have ATS systems that score resumes against job descriptions automatically. Candidates are flying blind. RACK gives you the same scoring intelligence — so you know before you apply whether your resume is a fit, and exactly which version to send.
Resume-Aware Candidate Kernel — the intelligent core that connects the right resume to the right job, every time.
This software is proprietary and confidential. See LICENSE for details. Unauthorized use, copying, or distribution is strictly prohibited.