Built Resume Intelligence with one rule: no backend, no database, resume never leaves the browser.
How it works:
PDF parsed 100% client-side via Web Workers (pdfjs-dist)
BYOK — bring your own Gemini/Groq/Mistral/OpenAI key, browser calls the provider directly
Strict JSON schema forces consistent AI output across providers: score, flagged lines, missing keywords, bullet rewrites
Bounding-box PDF inspector — click a flagged issue, it scrolls + highlights the exact line
Past scans stored in IndexedDB, not a server
Full build breakdown + copy-paste master prompt: https://dhaval-prajapati.onrender.com/blog/how-i-built-ats-resume-analyzer-blog
Try it free: https://ats-resume-analysis-sigma.vercel.app/
Top comments (0)