Rezumate helps job seekers upload a resume, paste a job description, get instant ATS-focused matching, rewrite weak bullet points, and export an ATS-safe PDF—processed entirely locally on their iPhone.
Resumes contain highly sensitive personal information, including names, phone numbers, email addresses, and full career history. Standard resume builders upload this data to third-party cloud servers.
Rezumate processes everything in-memory and stores history in secure, local sandboxed storage on the device. All AI tasks are handled on-device using local models, ensuring complete data privacy.
ios/- Native SwiftUI iOS app, built for App Store release.web/- Next.js marketing website containing waitlist, privacy, and support pages.
Open the Xcode project:
open ios/RezumateNative.xcodeprojThe app runs 100% locally and includes:
- Local Document Parsers: In-memory parsing of PDFs (
PDFKit) and Word Documents (DocxTextExtractorZIP XML scanner). - Deterministic ATS Scoring: Local calculations for keyword match, structure quality, and formatting warning checks.
- On-Device AI Rewrites: Powered by a quantized Llama 3.2 1B Instruct model running directly on Apple's Neural Engine.
- Local Storage: Saved variants and scoring history are persisted locally in secure, sandboxed storage via JSON structures.
- On-Device Export: Generates clean, selectable, single-column ATS-safe PDFs directly from local SwiftUI layout bounds.
Note: In debug builds, you can click "Use Local Dev Session" to bypass Apple authentication and run the app completely offline immediately.
cd web
npm install
npm run devThe marketing site serves as the landing page and waitlist capture. It can be deployed statically to Vercel, Netlify, or similar hosts.