CareerPath

AI résumé tailoring, interview prep and application tracking

careerpath-cv-tn.vercel.app

CareerPath rewrites your résumé and cover letter for one specific job. You paste the posting and get tailored documents plus a match score showing which of its requirements you meet and which you don't; from there you track the application on a board, practise against interview questions generated from that same posting including a live voice mock, and work through a day-by-day plan for the skills you're missing. I built the whole stack — Next.js App Router and Supabase with row-level security, an AI pipeline that falls back across Groq, Gemini and Anthropic behind one JSON-schema validator, a WYSIWYG editor that exports PDFs through headless Chromium at the layout shown on screen, Stripe and Dodo billing, a Chrome extension for clipping postings, and three locales including RTL Arabic. The hard part was regeneration: once someone has hand-edited a generated résumé, a fresh draft can't just replace it, so entries are matched by identity and the merge draws an explicit line between what the model owns (the writing) and what the user owns (their edits, per-entry dates, and the sections the generation schema can't express).