Generate and polish a resume locally without sending data to a server
Score an existing resume against a job description with AI feedback
Draft cover letters and translate resume text between Chinese and English
Run mock interview practice on a Mac or Windows desktop
Building from source needs Node 20, pnpm 9, a Rust toolchain, and Tauri 2 tooling; prebuilt installers are Windows-only with macOS still in progress.
JobPilot is a local-first AI job-hunting desktop app, mainly aimed at Chinese-speaking job seekers. The README describes it as a workstation that walks you through writing and polishing a resume, then practising for interviews, all on your own machine without needing to set up any server. You download an installer from GitHub Releases, double-click, and it runs. The original feature set, inherited from two earlier projects called JadeAI and RoleRover, covers the resume side of things. There is a drag-and-drop editor with inline editing and auto-save, more than 50 templates in styles like classic, modern, minimal, and "ATS-friendly" (meaning formats that automated recruiting systems can parse), and an AI assistant that can generate a resume, rewrite content, score it against a job description, draft cover letters, and translate or polish text. Existing resumes can be parsed from PDF or image files. Output can be exported to PDF, PNG, or Word, and you can also generate a shareable link or an AI-made LinkedIn-style headshot. Interface text is available in both Chinese and English. The JobPilot fork adds a Tauri-based native desktop wrapper, which the README says is lighter and faster than a typical web-shell desktop app and includes Apple Silicon Mac support in addition to Windows. Import now handles JSON, Markdown, PDF, and images (PNG, JPG, WebP), with scanned PDFs and image files routed through a multimodal model for parsing. There is a new Markdown editor with a toolbar for bold, italic, code, lists, and links, and a multi-line text-area component meant for long fields like work history. The app can check for and install its own updates and lets you refresh the list of available AI models manually. Planned next are LinkedIn-style certificate photos and resume version history with comparison and rollback. To build from source you need Node.js 20 or newer, pnpm 9, a stable Rust toolchain, and the Tauri 2 Windows tooling. The technology stack is Next.js 16 with React 19, Tauri 2, TypeScript 5, Tailwind CSS 4, shadcn/ui components, Zustand for state, and the Vercel AI SDK for model calls. The project is released under Apache 2.0 and credits JadeAI and RoleRover as the upstream codebases. Installers are currently Windows only, with a macOS build listed as in progress.
Generated 2026-05-22 · Model: sonnet-4-6 · Verify against the repo before relying on details.